COMPRESSION OF FOODS DURING VACUUM FREEZE DEHYDRATION
by
SEID-HOSSEIN EMAMI
B.S., University of California, Berkeley
(1974)
S.M., Massachusetts Institute of Technology
(1976)
SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE
DEGREE OF
DOCTOR OF PHILOSOPHY
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
August, 1979
...
...
Department of Nutrition and Food Science
Signature of Author -..
August
Certified
.....
by
10, 1979
.......
Thesis Supervisor
Accepted
by ...........
..............
ARCHiVESChairman,
MASSACHUSETTS
INSTITUTE
OFTECHNOLOGY
OCT 1 6 Uli
LIBRARES
.........
...
epartmental Committee on
Graduate Studies
COMPRESSION OF FOODS DURING VACUUM FREEZE DEHYDRATION
by
SEID-HOSSEIN EMAMI
Submitted to the Department of Nutrition and Food Science
on August 10, 1979 in partial fulfillment of the requirements
for the Degree of Doctor of Philosophy
ABSTRACT
Reversible reduction in volume of freeze-dried
food so that its dry density approaches that of fresh
product will result in significant efficiencies, especially
in the areas of package material requirements, storage, and
transportation.
The technique used in this study is simultaneous
compression of the food material by mechanical means during
the "normal" course of freeze-drying process to produce
low-moisture, high density product.
Compression of foods during freeze drying can be
considered as an efficient method, since it minimizes the
number of processing steps (i.e., initial overdrying,humidification and final redrying that are required in currently
practiced methods).
Investigation of the influence of several process
parameters on the degree of compression showed that the applied
compressive pressure is the main factor determining the final
degree of compression. It was noted that increasing the
applied pressure also acted to accelerate the drying process,
this being attributed to improved thermal conductivity and
heat transfer through the already dried, compressed layer.
As expected, increasing the level of heat input
accelerates the freeze-drying process. However, within
limits it did not have any apparent effect on the final
degree of compression.
The size of the sample cell holders used and porosity
of the sample holder material used in this study did not seem
to have any significant effects on drying and compression
behavior of the product.
This study has included an investigation of the
compression behavior of freeze-dried green beans rewetted
to moisture levels between 6% and 45% (dry solid basis) at
temperatures of -250 C to +240 C, using the Instron Universal
Testing Machine. The results showed that both moisture
and temperature strongly affect the mechanical properties
3
of the food. Within the range of moisture and temperature
investigated, the effect of increasing either or both of
these variables
can be described
as a decrease
in the
relative rigidity of the food structure, this being indicated
by the level of the stresses that are required to reach a
given strain and variation of the values of the modulus.
It was further shown that the effects of moisture
and temperature are inversely interactive so that there
exist specific combinations of moisture content and temperature which correspond to common values for mechanical properties and result in a similar compression response under
stress. This indicates that given either a value of
moisture and temperature or a value of a pertinent mechanical property, itis possible to predict and construct a
stress-strain relationship for freeze-dried green beans.
From evaluation of literature information on moisture
and temperature gradients that exist in the dry layer during
freeze drying, it was shown that the resulting values of
mechanical properties are suitable for compression during
freeze drying. These results indicated that the same
factors (moisture content and temperature) which govern the
compressive behavior of food in the objective testing method
using the Instron Universal Testing Machine also govern
the compression behavior during freeze drying.
The correlations of compressive behavior obtained
from the Instron studies also were used to evaluate the
results obtained in compression during freeze drying. Based
on these correlations, a simple method was developed for
the prediction of the compression behavior of the foods
during freeze drying.
Thesis Supervisors:
Dr. James M. Flink, Associate Professor of Food
Engineering
Dr. Marcus Karel, Professor of Food Engineering
4
ACKNOWLEDGEMENTS
While realizing the futility of adequately expressing my gratitude, I will nonetheless make an effort
to thank all those who have made this work possible.
First, to Professors James M. Flink and Marcus
Karel for their help, guidance and advice throughout this
study and my education at M.I.T.
I would especially like
to express my sincere gratitude to Professor James M.
Flink for his constant assistance and helpful advice
which he so generously devoted toward this project.
I would also like to extend my deepest appreciation
to the members of my doctoral committee, Professors D. I. C.
Wang, A. J. Sinskey, and especially Professor C. Rha for
their time, effort and guidance throughout this study.
Thanks are also extended to James Hawkes for his
continuous help and drawing the figures, and to Joanne
Klotz for typing the manuscript of this thesis.
I would
like to acknowledge
all the members
of the
Food Engineering group, especially Steve Haralampu and
Eddie To for their friendly assistance during the completion of my work.
This study is dedicated to my dearest parents, for
without them lots of opportunities I have today would seem
impossible.
This project was supported by the U.S. Army Natick.
Research and Development Command, Contract No. DAAKQ3-75-C-0Q39.
5
CONTENTS
Page
ABSTRACT
. . . . . . . . . . . . . . . . . . . . . .
2
ACKNOWLEDGEMENTS
. . . . . . . . . . . . . . . . . . .*
LIST
OF
FIGURES
. . . . . . . . . . . . . . . . . . .
LIST
OF
TABLES
1.
INTRODUCTION
2.
LITERATURE
14
. . . . . . . . . . . . . . . . . .
15
SURVEY
. . . . . . . . . . . . . . . ·
.
19
. . . . . . . . . . . . . . .
19
Freeze
2.2
State of Water and Solutes in Frozen
2.3
Drying
Freeze-Dried
2.3.2
21
.
. . . . .
. . . . . . . . . .
22
Evidence for a Diffused Sublimation
Front
2.6
20
Evidence for a Sharp Sublimation
Front
2.5
. . . . . . . . . ·,
Systems
State of Water during Freeze Drying .
2.3.1
2.4
10
. . . . . . . . . . . . . . . . . . . .
2.1
and
4
24
. . . . . . . . . . . . . . .
Compression; A Rheological Approach . .
2.4.1
Compressive
2.4.2
Degree
of
27
. . . . . . . . .
27
Stress
. .
Compression
. . . . . . . .
28
2.4.2.1
Compressive
Strain
. . . . . .
29
2.4.2.2
Compression
Ratio
. . . . . ·
.
30
.
31
. . . ,.
33
Effect of Moisture and Temperature on
Mechanical (Rheological) Properties of
Food (and Other Polymeric Compounds) . . . .
35
2.4.3
Compression
2.4.4
Compression Behavior Analysis
Behavior
. . . . . . . .
Methods of Compression and Debulking Dehydrated
Foodstuffs
. . . . . . . . . .
40
6
Page
2.7
Alternative Approaches for Production
of
. . . .
Foods
Freeze-Dried/Compressed
.
42
..... ... 42
Freeze
Drying
2.7.1
Limited
. . . . . . . . .
2.7.2
Microwave
Technique
2.7.3
Simultaneous Freeze Drying and
Compression
2.7.3.1
. . . . . . . . . . . . .
47
. . . . .
48
Behavior
Compression
2.7.3.2 Effect of Sample Thickness .
51
2.7.3.3 Final Moisture Contents
53
. . .
2.7.3.4 Rehydration Behavior .
...
. . .
2.7.3.5 Organoleptic Evaluation
2.8
3.
Moisture-Temperature Gradients in the
Dry Layer during Freeze Drying; Their
Effects on Mechanical Properties of the
Dry Layer during Freeze Drying . . . . ..
EXPERIMENTAL
3.1
46
. . . . . . . . . . . . . .
53
55
..
55
. .
63
. . .
63
Equipment for Compression during Freeze
Drying.
.
. . . . . . . . . . .
.
3.1.1
Hydraulic Compression System ....
3.1.2
Vapor
3.1.3
Sample
Cell
3.1.4
Sample
Loading
3.1.5
Program for Compression during
Freeze
. . . .
System
Trapping
.
.
. . . . . . . ...
Holder
.
..
.
.
.
. .
.
. . . . ..
Experiments
Drying
63
65
66
68
69
3.1.5.1 Compressive Pressure or
Loading
. . . . .
....
3.1.5.2 Heater Controls/Energy Input .
69
71
3.1.5.3 Sample Holder Size and
Porosity.
...
.
..
71
7
Page
3.1.6
3.2
Terminology for Compression during
Freeze Drying Experiments ......
72
Materials and Methods Used in Evaluation
of the Rheological and Mechanical Properties
of Green Beans as a Function of Moisture and
Temperature(InstronStudies). . . . . ...
3.2.1
Freeze Drying
3.2.2
Rehumidification
3.2.3
Compression with the Instron Universal
Testing
4.
RESULTS
4.1
AND
Machine
DISCUSSION
.
73
. . . . . . . . . . .
73
.
. . . . . . . . .
. . . . . . . . . .
..
4.1.5
. . . . . .
....
78
. . . .
.....
.
87
Relations of Extents of Drying and
Compression Relative to the Applied
Compressive
4.1.4
.
Dependence of the Drying Rates on
Sample Loading (Applied Compressive
Pressure)
4.1.3
78
Dependence of the Degree of Compression on Sample Loading; Compressive Pressure-Strain
Relationship
4.1.2
75
78
Compression during Freeze Drying Results
4.1.1
4.2
.........
73
Pressure
.
. . . . . . .
89
Effect of Heat Input on the Compression and Drying Behavior .....
91
Effect of Size and Porosity on Com. . .
pression and Drying Behavior .
95
Compression Behavior of Green Beans;
Moisture and Temperature Dependencies as
Evaluated with the Instron Universal Testing
. . . . . . . . . . . . . . .
96
Effect of Moisture and Temperature
on Compressibility of Freeze-Dried
..........
Green Beans .
98
Machine
4.2.1
4.2.2
Effect of Cross-Head Speed or Strain
Rate on Compressibility . . . .
105
8
Page
4.2.3
Effect of Moisture and Temperature
on
4.2.4
4.3
Mechanical
Properties
.
. . . . .
Moisture-Temperature Dependencies;
Common Mechanical Properties and
Similar Compression Behavior
(Stress-Strain Relationship) .....
Correlation of Compression Behavior during
Freeze Drying to Compression Behavior with
the Instron Universal Testing Machine . . . .
4.3.1
114
Pressure
.
. . . .
..
123
Development of Means to Predict
Compression Behavior during Freeze
Drying; Compressive Pressure-Strain
Relationship
4.3.4
114
Mechanism of Compression during
Freeze Drying at Constant Applied
Compressive
4.3.3
110
Moisture-Temperature Relationships
and Mechanical Properties of the Dry
Layer during Freeze Drying . . . . .
4.3.2
105
. . . . . . . . . . ...
123
Evaluation of the Results Obtained
in Compression during Freeze Drying
Experiments Relative to the Parameters Tested Based on the Results
Obtained in Instron Studies .130
4.3.4.1 Effect of the Applied Compressive Pressure on the
Drying Rates and the Degree
of
Compression
. . . . . . . .
130
4.3.4.2 Effect of Increasing Levels
of Heat Input on Drying
Rates and the Degree of
Compression
. . . . . . .
134
4.3.4.3 Effect of Size and Porosity
of the Sample Holders on
Drying and Compression
Behavior
4.4
General
Discussion
.
.
..
. . . . . . . . . . . .
138
140
9
Page
5.
CONCLUSIONS
6.
FUTURE
REFERENCES
.
RESEARCH
.
.
.
................
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
146
.
.
.
.
.
.
.
.
.
.
.
.
.
.
149
.
.
153
10
FIGURES
Figure No.
1-1
Page
Methods of preparation of freeze-dried/
products
compressed
2-1
2-2
2-3
2-4
32
A typical stress-strain response for
freeze-dried, rehumidifed green beans under
...
.
a uniaxial compression. ..
34
Methods of preparation of freeze-dried/
.......
compressed products .
43
Schematic of spring compression system
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .50
Dependence of compression ratio on sample
thickness
2-7
for
green
. . . . . . . . .
beans
Effect of mode of water addition
beans
2-11
..
. . . .
.
56
. . . . . .
cubes
. . . . . . . . . . . . . . . . .
Models of moisture gradients present
during freeze drying (from Karel, 1975)
.
.
57
59
Local moisture contents in the dry layer
during freeze drying as a function of
difference between local temperature and
ice-front temperature (from Aguillera
and
3-1
54
Effect of the degree of compression on
rehydration of freeze-dried compressed
beef
2-10
. . .
Effect of the degree of compression on
rehydration of freeze-dried, compressed
green
2-9
52
on re-
hydration of freeze-dried green beans
2-8
49
Effect of compressive pressure on compression ratios of green beans for spring
system
2-6
18
The 4-element model; the basic combination
of Maxwell model and Kelvin model . . . . .
(loaded)
2-5
. . . . . . . . . . . .
Flink,
1974)
.
. . . . . . . . .
.
61
Schematic of the hydraulic compression
system
. . . . . . . . . .
64
11
Page
Figure No.
3-2
Schematic of the sample cell holder for
compression during freeze drying
. . . . . . . . . . . . . . . .
experiments
3-3
Schematic of the sample cell holder for
compression tests using the Instron
Testing
Universal
4-1
Machine
.
76
.
.. .
Effect of compressive prssure on degree
of compression
(strain) of green beans at
81
different levels of heat input .......
4-2
Effect of compressive pressure on strain
(corrected for initial moisture content)
of green beans at different levels of
heat
4-3
input
. . . . . . . . . . . . .
.
4-5
4-8
of
green
beans
...
. . . . . . .
pressure
. . . . . . . . . ...
Effect of heat input, cell wall porosity,
and size on drying rates of green beans . .
88
90
93
Percent reduction in volume vs. percent
water removed during freeze drying of
green beans at different levels of heat
. . . . . . . . . . . . . . . . ...
94
Percent reduction in volume vs. percent
water removed during freeze drying of
green beans at different sample holder
sizes
4-10
86
Percent reduction in volume vs. percent
water removed during freeze drying of
green beans at different levels of com-
input
4-9
.
85
Effect of compressive pressure on drying
pressive
4-7
.
.
Compression behavior of green beans as
function of time during freeze drying at
different levels of compressive pressure
rates
4-6
83
Effect of compressive pressure on compression ratio (corrected for initial
moisture content) of green beans at dif-
ferent levels of heat input . . . .
4-4
67
and
wall
porosities
. . . . . . . . .
97
Effect of moisture content on stress to
achieve given strains at +24°C (Instron
tests)
. . . . . . . . . ..
. .
.. . . . .
99
12
Page
Figure No.
4-11
Effect of moisture content on stress to
achieve given strains at +150C (Instron
tests)
4-12
4-15
4-16
4-18
4-19
4-20
.
.
.
.
.
.
. .
.
.
.
.
.
.
. . . .
. . . . . . . . . . . . . . . . . . .
100
101
102
Effect of moisture content on stress to
reach a strain of 0.75 at different
temperatures(Instrontests) . . . . . ...
104
Effect of moisture content on stress to
achieve given strains at different crosshead speeds (+24°C, Instron tests) . .
106
(E)
Effect of moisture content (and temperature)
on tangent modulus calculated at strain
(Instron
tests)
. . . . . . .107
Effect of moisture content (and temperature) on tangent modulus calculated at
strain () = 0.30 (Instron tests) . . . .
108
Effect of moisture content (and temperature) on secant modulus calculated at
stress (a) = 21 psi (Instron tests) . . .
109
Moisture-temperature relationship with
equivalent compression behavior at
specified values of tangent modulus (at
= 0.65) obtained in Instron tests . .
111
Moisture-temperature relationship with
equivalent compression behavior at
specified values of tangent modulus (at
= 0.30) obtained
4-21
. .
. . . . . . . . . .
= 0.65
4-17
.
Effect of moisture content on stress to
achieve given strains at -25°C (Instron
tests)
4-14
. . . . . .
Effect of moisture content on stress to
achieve given strains at -10°C (Instron
tests)
4-13
.
in Instron
112
tests
Moisture-temperature relationship with
equivalent compression behavior at specified
values of secant modulus (at a = 21 psi)
obtained
in
Instron
tests
.
. . . .
.
113
13
Page
Figure No.
4-22
Estimated local moisture contents in the
dry layer during freeze drying of green
beans as a function of difference between
local temperature and ice-front tempera-ture
4-23
4-24
of
-25C
. . . . . .
116
Estimated values of tangent modulus (at
= 0.65) in the dry layer during freeze
drying of green beans as a function of
distance in the dry layer . . . . .
118
Estimated range of tangent modulus (at
= 0.30) that exist in the dry layer dur-
ing freeze drying of green beans . . . . . .
4-25
4-26
Estimated range of secant modulus (at
a = 21 psi) that exist in the dry layer
during freeze drying of green beans
Behavior of modulus with respect to time
and location during freeze drying at constant
4-27
temperature
ice-front
122
. . . . . .
Representation of the equivalent moisture
content simulating compression during
freeze drying data at Instron test temperature
4-28
120
10
of
+15C
. . . . . . . . . . . . . . .
Representation of the equivalent moisture
contents simulating compression during
freeze drying data at Instron test tempera-
turesof +240 C and -10C . . . .
4-29
Errors in estimation of compressive pressures to reach given strains in compression during freeze drying caused by
+1% variation in equivalent moisture
content (obtained from Instron tests
at +24°C)
4-30
4-31
127
.
.. . .
.
132
Effect of increasing the surface temperature (Ts) during freeze drying on the
range of modulus in the dry layer at
.
constant ice-front temperature .
137
Compressibility behavior of different
locations in the dry layer during freeze
drying
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
TABLES
Page
Table No.
3-1
Program for Compression during Freeze
Drying
3-2
4-1
Experiments
. . . . . . . .
..
Relative humidities used for rehumidification of freeze-dried green beans for
studies of mechanical properties (from
.
.
....
Labuza et al., 1976) .
4-3
Specific combinations of moisture and
temperature with corresponding values of
the mechanical properties which simulate
compression during freeze drying data . . .
79
128
Variations in the stresses to reach
given strains caused by +1% variation in
the estimation of the equivalent moisture
content obtained from Instron studies at
+24C
4-4
74
Results of compression/freeze drying
experiments conducted with green beans at
various freeze drying conditions . . . . ..
4-2
70
. . . . . . . ..
. . . . . ...
The overall effects of parameters tested
in compression during freeze drying on
final degree of compression and drying
time of green beans ..........
131
133
15
1.
INTRODUCTION
Freeze drying has long been recognized as the method
of dehydration giving products with high quality and retained
Freeze-dried foods can be quickly
nutritional value.
rehydrated and when rehydrated they closely resemble the
food prior to dehydration.
When foods undergo freeze drying,
however, there is very little volume reduction from the
original food form.
While the retention of physical shape has long been
viewed as desirable since it indicated no structural change
in the product, it does result in an open and porous structure
which gives a very light, fragile product of low packaging
The disadvantages related to low packaging density
density.
can be summerized
as:
·
The open structure is susceptible to oxidation
and moisture pickup, and thus special packaging
is often required.
·
Large package volumes are required to enclose
small product weights.
*
Readily friable products produce fines during
handling and transportation which is essentially
a product wastage.
*
Large storage space is required.
Under most situations, it becomes extremely advantageous to compact dehydrated foods by reducing or
eliminating the void volumes which exist between individual
pieces and the internal void space which was occupied by
water prior to dehydration.
For such compacted or compressed
16
foods, there are greatly reduced packaging, handling,
transportation, and storage requirements which gives a significant potential economic savings.
The requirements made
on the compressed food products are that upon rehydration
they regain their original geometry and resemble their noncompressed counterparts in color, odor, flavor, texture, and
appearance.
To avoid fragmentation and irreversible structural
changes the product must be in a compressible state during
the compression process.
The currently practiced method for producing
freeze-dried and compressed foods has been described by
Rahman et al.
(1969).
In short, the product
is pretreated
as required, blast frozen and freeze dried to a final
moisture content of less than 2%.
The freeze-dried product
is then made compressible by either thermal treatment
(Rahman et al., 1970) or by addition of low-molecular-weight
agents such as water (Hamdy, 1962; Ishler, 1965) or glycerine,
glycerol, propylene glycol (Brockman, 1966; Rahman et al., 1969).
Water is considered to be the most desirable agent since it
imparts no foreign flavor to the rehydrated product, and
the current industrial practices for addition of moisture
to freeze-dried products prior to compression include steaming
and water misting (Rahman, personal communication).
In
both rewetting processes, the freeze-dried foods are brought
to 5% to 15% moisture.
17
Compression is accomplished uniaxially at pressures
between 250 to 2000 psi after which the product is redried
to a moisture content of less than 2% and packaged.
The overdrying, rewetting, and subsequent redrying
of the product presents a major inefficiency in the process,
and studies on other methods for preparing freeze-dried/
compressed foods have been initiated such as "limited
freeze drying" and "partial freeze drying and microwave
treatment."
The studies to be reported on herein have considered
an approach where compression and drying of the food material
to a low-moisture, high-density product is accomplished
simultaneously and in a single-step process.
In this process,when food is under stress during
freeze drying, it undergoes a reduction in volume as well
as weight which can be later recovered by rehydration of
the food.
Figure 1-1 shows the different methods that have
been used to produce freeze-dried/compressed foods.
Included in this, there are studies of the mechanical properties of the food which relate to the compression
behavior during the freeze-drying process.
18
U
z
0
z
0
t
l
a.
W-_ a
-rJ
2 llJ
o
ijU-
z
0
4
;,
t
z
t
--0
aO
W
ro
a)
cJ
Z
(E
)
o
1)
0
Q)
In
. -J
0
z
(D
z
-O
(CD
w
N
·-H
rd
L
4z
W."
N
zZ
LL
J
U)
w
(N
C
NaZ
LCZ
Q.U.
uc r0
w
w
o4o(1)
4-
_
W
J
o0
a4
(O
04
i
to
I-N
nW
t cn
z
z
0
r--
-c
ww
0. cn
r--i
.11
19
2.
2.1
LITERATURE SURVEY
Freeze Drying
Vacuum freeze drying is a method of dehydration in
which water is removed from foods by transfer from the
solid state (ice) to the gaseous state (water vapor).
This
sublimation process can be accomplished by reducing the
temperature and vapor pressure of water below the triple
point
(about 0°C and 4.5 torr) and by supplying
energy of sublimation (about 600 cal/gm).
the
This can be
achieved by (1) radiation to the outer dry layer of the
drying food, (2) conduction through the frozen layer, or
(3) internal heat generation (microwaves). Lastly, it is
necessary to remove the water vapor generated by sublimation
with a moisture sink (King, 1971).
Advantages of this technique include the low processing temperature, the relative absence of liquid water and
rapid transition of any local region of the material from
a fully hydrated to a nearly completely dehydrated state.
This rapid transition minimizes the extent of various degradative reactions which often occur during drying.
The low
temperatures involved also help to minimize these reactions.
The absence of the liquid state further helps to minimize
degradative reactions and prevents surface concentration
of solutes leading to poor rehydration ability (King, 1971).
Despite the fact that freeze drying gives dehydrated
products of very high quality and high retained nutritional
20
value, it involves high initial capital investment, high
processing costs, for some products it requires special
treatment (cutting, slicing, etc.) prior to drying to improve
dehydration, and results in products of very low density
(Longmore,
2.2
1973).
State of Water and Solutes in Frozen
and Freeze-Dried Systems
The most important change that occurs during freez-
ing is the appearance of at least one new phase (i.e., conversion of water into ice).
According to Mackenzie (1965),
other changes include:
*
Dehydration of the insoluble solid material
*
Coalescence of droplets of immiscible liquids
*
Crystallization of some solutes
*
Formation of amorphous precipitates of other
solutes
*
Exclusion of dissolved gases
·
Formation of concentrated solute glasses
*
Disruption of molecular complexes (lipo-proteins)
The ideal frozen state would show a sharp and discrete dividing surface between a region which consists of
ice crystals and a region which consists of concentrated
amorphous solution.
However, there is substantial evidence
that of the total water present in food, a portion is strongly
bound to individual sites of the solute components, while
an additional amount is bound less firmly, but still is not
available as solvent for various food components (Karel, 1974).
21
It has been established that upon cooling water-containing
foods to low temperatures, well below the freezing point of
the food, a portion of the total water remains unfrozen.
This "unfreezable water," which is estimated to be in the
range of 0.2 - 0.4 g per g food solids is less free than
liquid water but has greater mobility than ice.
The portion
of water which is held more strongly than "unfreezable water"
is called "monolayer value" and is assumed to be bound
to specific sites in foods.
Foods contain many polar sites
including, for instance, the hydroxyl groups of polysaccharides, carbonyl and amino groups of proteins, and
others to which water can be held by hydrogen bonding, iondipole bonds or by other strong interactions (Karel, 1974).
The monolayer value is well below the unfreezable water
content, it being estimated to be 20%-50% of the nonfreezing fraction of water (Duckworth, 1972).
2.3
State of Water during Freeze Drying
During freeze drying, the ice region retreats inward
as a front, leaving the dried matrix behind.
The resultant
structure consists of pores (the location of former ice
crystals) passing through a dry matrix of insoluble components
and precipitated compounds which were originally in solution.
It should be noted that if during freeze drying the temperature and moisture content in any region surpasses a critical
level, mobility in the concentrated aqueous solution may
be sufficiently high to result in flow and loss of the
22
original structure (melting or collapse) and loss of the
separation of solutes which existed immediately following
freezing (Karel, 1975).
Most analyses of freeze drying kinetics have been
based on an assumption of a sharp and discrete dividing
surface between a region which is fully hydrated and frozen,
and a region which is nearly completely dry.
However, in
recent years there has been considerable discussion regarding
the degree of sharpness of this surface when it is retreating
as a frozen front during freeze drying.
2.3.1
Evidence for a Sharp Sublimation Front
King (1971) explains that a very sharp front would
result if sublimation occurred from a very thin zone near
the surface of the frozen region, and if the sublimation
removed essentially all of the initial moisture from the
remaining solid material immediately after the passage of
the frozen front from that region.
Many investigators have reported cutting open
partially freeze-dried specimens and observing a distinct
and relatively sharp demarcation between the still frozen
zone and dry zone of uniform coloration.
These include
Hardin (1965) who studied freeze-dried beef, Margarities and
King (1969) who studied freeze drying turkey meat and
Beke (1969) who studied freeze drying of pork.
Additional evidence is obtained by Hatcher (1964)
who used gamma ray attenuation measurements to monitor
23
the profile of moisture content across relatively thick
(2 inches) slabs of beef as a function of time during
freeze drying.
Hatcher reports that visual inspection of
partially dried specimens revealed the ice front to be
Since his gamma ray beam diameter was 3/16 inch,
planar.
he concluded that the ice front was sharper than could be
detected by the beam.
He also found that the residual
moisture in the dry zone was so low that it did not affect
the gamma ray count rate.
MacKenzie (1966) constructed a "freeze drying
microscope" in which he could observe freeze drying as it
occurred in various transparent frozen liquid systems.
An
extremely sharp sublimation front was observed, although
the front, while sharp, did not necessarily remain planar.
There should be a finite partial pressure of water vapor in
the gas within the dry layer, since the water vapor generated
by sublimation must escape across the dry layer.
Therefore
it should be expected that a finite amount of sorbed water
will be present in the dry layer corresponding to at least
the amount that would be predicted by equilibrium sorption
isotherm.
Sandall (1966) made calculations for typical
conditions of freeze drying of turkey breast meat using
the sorption isotherm for the same material that was measured
by King et al. (1968). For heating from the outer dry
surface,he concluded that equilibrium between the dry
layer and the escaping water vapor would give an average
24
moisture content within the dry layer equal to 3 percent
of the initial water, or about 1.5 percent moisture content
on a solid basis.
However, this surprisingly low value
seems to result from the increasing temperatures across the
dry layer toward the outer surface as well as from the
decreasing water vapor partial pressure (King, 1971).
2.3.2
Evidence for a Diffused Sublimation Front
Meffert (1963) measured temperature profiles during
freeze drying of rutabaga.
From the temperature and thermal
conductivities he inferred from them, it was concluded that
some 30 to 40 percent of the initial water content was left
behind by the retreating ice front.
This remaining water
had to be removed by secondary drying or desorption.
In
1965 Meffert determined moisture contents in layers of
suede, and found basically the characteristic two-zone phenomenon,
a dried layer and an ice core.
Also, clearly distinguishable
was the existence of a moisture gradient through the dried
layer that varied from around 3.0 g H 2 0/100 g solids near the
outer surface to 33.0 g H 2 0/100 g solids near the interface
when the ice core had receded half way into a 3.6 cm sample
dried from both sides.
Brajnikov et al. (1969) interpreted that the results
of some experiments demonstrated the presence of a diffuse
sublimation front.
He took color photographs of cut
specimens of beef which had been freeze-dried once, rehydrated with a solution of CoCl2 and then partially
25
freeze-dried again.
Divalent cobalt has the property of chang-
ing color from pink to blue depending on water activity of
its surroundings.
Brajnikov's experiments with CoC12 in-
dicated 'the presence of a transition zone; however,
King (1971) has noted that the cobalt salt can lower the
freezing point which could cause the presence of some liquid
at the sublimation front, even at -200 C.
Luikov (1969) observed that sublimation and ablimation
(desublimation from vapor state to solid state) occur simultaneously at the sublimation front of an ice sphere.
The
tendency for ablimation to occur depends upon the degree
of water vapor saturation of surrounding gas phase.
Ablima-
tion occurred to a greater extent as the surrounding gas
became more nearly saturated in water vapor.
The degree of sharpness of the ice front during
freeze drying reflects the extent of freeze drying and the
condition(s) existing during freeze drying.
Important
parameters which seem directly related to broadening of the
ice front are pressure, and temperature difference between
local temperature in the "dry" layer and ice front temperature.
Luyet (1962) in describing drying phenomena includes
a "pseudo freeze drying" or "secondary drying" stage which
accounts for the withdrawal of the non-frozen water.
MacKenzie and Luyet (1964) observed with a freeze drying
microscope that the secondary drying of the outermost
26
portions of a sample proceeds simultaneously with. sublimation
of ice from the sample interior.
Bralsford (1967) described secondary drying as
taking place at the interface between the dried and frozen
region.
In his description the interface had a finite
thickness which increased as the drying front receded toward
the interior of the specimen.
He postulated the existence
of a "diffusion zone" whose broadening as the zone recedes
into the sample was due to solute migration which gives more
and more ice being melted.
He felt that at higher chamber
pressures more liquid water is present in the ice core which
results in some liquid diffusion into the dry layer.
Aguilera and Flink (1974) calculated moisture profiles
from temperature measurements obtained during freeze drying.
Their calculation was based on a relationship originally
noted by Gentzler and Schmidt (1973) between the bound water
content at a position in the dry layer (WB) and temperature
difference (ATB ) between that position in the dry layer
(desorption temperature) and the sublimation temperature of
ice during the freeze drying.
When the mathematical
expression is based on higher desorption temperatures for
bound water as compared with the sublimation temperature for
frozen water,
WB
=e
B
it can be expressed
(aATB+b)
as:
27
where a and b are constants dependent on each particular
product.
Aguilera and Flink (1974) indicated that combining
the measured temperature profiles with the above relationship results in higher moisture contents being calculated
at a given distance from the ice front as drying proceeds.
Their results show that at a heater temperature of 128°C and
a surface temperature of 56°C, the diffusion zone was about
2 mm when the ice front had retreated 5mm (after 2.5 hr),
while it had expanded to 5 mm thick when the ice had retreated
about 15 mm (after 9 hr).
While quantified drying theories, which have assumed
a sharp and uniformly retreating ice front (URIF) which
leaves behind an essentially dry layer, do not seem to result
in large errors in predicting drying times for various
materials, it nevertheless seems clear that the existence of
a sharp sublimation front is an idealization (Karel, 1975).
2.4
Compression; A Rheological Approach
2.4.1
Compressive Stress
Stress has been defined (Mohsenin, 1968) as the
aggregates of all the tractions (force per unit area)
corresponding to all directions passing through a point in
the body.
Complete specification of stress requires the
use of at least six "components of stress."
Compressive
stress is one of the six components of stress.
It expresses
the linear, uniaxial application of force which causes
28
In this text the
compaction type volumetric change only.
usage of the term stress refers to compressive stress unless
specified.
2.4.2
Degree of Compression
Degree of compression represents the change in volume
of a body relative to its original, precompressed volume.
It can be expressed numerically in a number of ways, and
its numerical value will depend on the method of measuring
or defining volume (Emami et al., 1979).
A "displacement" type change considers only the
volume of individual sample pieces; no void volume of the
bulk sample is considered.
Thus, a "displacement" change
measure represents the true volume change of a sample
piece.
Since, in compression of freeze-dried foods, voids
that are produced by sublimation of ice crystals are reduced or eliminated, the maximum degree of compression of
an individual piece is related to its initial water content.
The technique used for determining the volume change of a
sample piece is similar to the technique for measuring loaf
volume of bread using rape seeds.
When the degree of com-
pression is measured by the change in initial and final
volumes of the sample pieces when orderly "packed" in the
sample holder, the void volumes between the food pieces
are also included in the measurement.
for when the food pieces are
This is also true
"randomly packed," for which
the initial volume is defined as the volume which is
29
occupied by the given sample when randomly packed in their
normal container or package.
The randomly packed volume is
much larger than when the sample is orderly packed in the
sample holder, due to larger void volumes between the sample.
Values given in literature for degrees of compression
are usually based on randomly packed volumes (Emami et
al., 1976).
2.4.2.1 Compressive Strain
Strain has been defined (Mohsenin, 1968) as the
relative change in linear dimension, or shape, or volume of
an element within a body.
Strain, like stress, can be
resolved into six components of normal and tangential
Compressive strain is one of the normal strains
strains.
referring to linear, uniaxial compaction type change in
volume only, and it indicates the percentage change in
volume which is given by:
= [(Vi-Vf)/Vi]
= strain
where:
V. = initial volume
1
Vf = final volume
Numerical values of strain increase from zero (no compression)
and have-higher values which approach 1 as the degree of
30
compression increases (i.e., final volume of the product
decreases).
Since compressive strain is unidirectional
(i.e, length and width do not vary), the change in height of
the sample can be used instead of change of volume (Emami
et al., 1979), such that,
£ =
where:
(hi-hf)/h
i
hi = initial height
hf = final height
In this text the use of the term strain refers to compressive
strain unless specified.
2.4.2.2 Compression Ratio
Compression ratio is defined as the ratio of the
initial sample volume.prior to compression to its final
volume (C.R. = Vi/V f) .
The compression ratio thus increases
from a value of 1 (no compression) and has higher values as
the degree of compression increases (i.e., volume of the final
product decreases).
Compression ratio indicates the factor
by which the volume of the sample has decreased due to
compression.
Like compressive strain, when the compression
is unidirectional (i.e., length and width do not vary),
height changes can be used to determine compression ratio
such that C.R. = hi/h
f.
1f
31
2.4.3
Compression Behavior
In all materials the deformation that occurs is
some function of the applied force.
For example, linear
elastic (Hookean) materials such as an ideal spring (Mohsenin,
1968), have a relation between stress and strain (in general
form) which is given by:
a =Es
where a is the applied stress,
is the resulting strain,
and E is "Young's modulus" (the slope of the straight line
relating stress and strain, which defines the stiffness or
rigidity of the given material).
In viscoelastic materials, the force deformation
characteristics are like elastic materials on short time
scales and are like viscous fluids on long time scales (a
combined solid-like and liquid-like behavior).
Materials of
this type are modeled by combinations of springs and dashpots, with the spring element in the mechanical model obeying
Hook's law while the dashpot represents a Newtonian liquid.
The two basic combinations of these elements, the Kelvin model
and Maxwell model, are shown in figure 2-1 (Mohasenin,
1968).
It is generally accepted that for small deformations
the rheological properties of solid foods can be described
by some sort of complex arrangement of springs and dashpots.
32
DASHPO
Fig. 2-1
SPRING
The 4-element model; the basic combination
of Maxwell model and Kelvin model.
33
For large, irreversible deformations in food, deviations
from elastic or viscoelastic behavior require the addition of fracture elements to the mechanical models (Drake,
1971; Peleg,
2.4.4
1976).
Compression Behavior Analysis
From the force-deformation measurements made with
an objective method such as an Instron Universal Testing
Machine, numerous quantifiable parameters may be calculated
to describe a material's texture.
For a uniaxial compression
at a fixed crosshead speed (i.e., constant strain rate) of
a material with an initial thickness of h
sectional
(
area of A0 , the stress
and fixed cross-
(a = force/A 0 ), and the strain
= (h0 -h)/h0 ), or compression ratio (C.R. = h0 /h) can be
obtained from the force-deformation data.
Figure 2-2 shows a typical stress-strain response for
freeze-dried, rehumidifed green beans compressed uniaxially
with the Instron Universal Testing Machine.
The definition
of the textural parameters shown in figure 2-2 is given by
Mohsenin
(1968) as follows:
Secant Modulus:
The ratio of stress to strain at
a given point on stress-strain
curve (a/s). It is an indication of
stiffness or rigidity.
Tangent Modulus:
The slope at a given point on
the stress-strain curve (ac/a).
It is also an indication of
stiffness or rigidity of a material.
34
Fig. 2-2
A typical stress-strain response for freezedried, rehumidifed green beans under a uniaxial
compression.
II
m
_J
Ix
n
"0
1.0
STRAIN e
COMPRSSION RATIO,
---
0.8
5C
I
- vI
35
Elastic deformation:
Recoverable deformation of a
material
Plastic deformation:
Permanent or unrecoverable
deformation of a material
2.5
Effect of Moisture and Temperature on
Mechanical (Rheological) Properties of
Foods (and Other Polymeric Compounds)
Adsorbedwater has been found to have pronounced
effects on mechanical properties of polymeric materials.
However, when the whole spectrum of water contents of the
material is considered, the effect of water varies significantly at different levels of adsorption.
At 0% water content, the material is found to be
quite rigid,and fragmentation will occur when the material
is subjected to stress.
At low water adsorption levels, the
adsorbate forms strong bonds (e.g., H-bonds, ion-dipole
bonds, etc.) to many specific adsorption sites (e.g.,
hydroxyl groups, carbonyl groups, amino groups, etc.) on
the material and does not have its original liquid form (i.e.,
it is present as "bound" water) (Karel, 1974).
At these levels
there is no significant effect of water addition on the
relative rigidity of the material (G. King, 1950).
At higher water adsorption levels, the relative
mobility of polymer chains and the volume they occupy
increase (G. King, 1950).
It is known that water and other
low-molecular-weight compounds (solvents) which can form
internal polar bonding or dissolve the polymer (Sakata
36
and Senjue (1975) reduce the local friction coefficient, and
increase the mobility of the polymeric segments.
Each
polymeric segment having adsorbate molecules in its vicinity
(as wll
as other polymeric segments) can be displaced in
translatory motion much more easily, thus lowering the
effective local viscosity (Ferry, 1961).
For the case of
most foods, this effect of moisture on relative rigidity
is reported for water contents in the range of 5-20 percent.
Compression of freeze-dried foods having moisture contents
in this range will result in a significant "plastic" deformation (i.e., no recovery after unloading), together with some
elastic recovery.
The plastic deformation results in
elimination or reduction of void volumes which were initially
occupied by water.
In the compression process, polymeric
segments initially separated by these void volumes now have
a greater degree of mobility and can be brought closer to
one another and be packed without significant breakage or
fragmentation.
This close packing will result in formation
of many polar bonds (e.g., H-bonds) between the polymeric
segments (Nissan, 1976) which is responsible for the nonrecoverable aspects of the deformation.
If the sample
is unloaded at the same moisture level used for loading, there
will be a time-dependent elastic recovery (i.e., relaxation),
which is due to breakage of some of the bonds which were
formed during loading (Nissan, 1976).
If immediately after
compression the sample is further dried (i.e., water is
37
removed), the relative rigidity of the polymeric segments
will increase again causing an indefinite deformation so
long as the sample is kept dry.
On the other hand, if
after unloading further water is added (i.e., rehydration),
the bonds formed during compression will break at a faster
rate, and water migrates back within the polymeric segments, separating them again.
The extent of water uptake,
which depends on the basic material structure and the strength
of the bonds formed during the compression process, will
determine the extent to which the sample approaches the
original shape and volume.
At high moisture levels the reduction in relative
rigidity with increasing moisture levels off (G. King,
1950), apprently since at these levels, liquid water is
present in a loose form within the polymeric segments.
Compression of the material at high water activities causes
collapse or flow which is accompanied by the loss of the
original solid structure of the matrix (MacKenzie and Luyet,
1971).
There have been very few studies on compressive
rheological properties of foods as a function of moisture
content and/or temperature.
Kapsalis et al. (1970) reported
significant changes in textural properties of freeze-dried,
precooked meat when rehumidified to water activities between
0 to 1.0 at room temperature.
According to their data, at
the range of moisture contents that have been found to be
38
suitable for compression of this product (11-12
g H2 0/100 g
solids), (Pilsworth, Jr. and Hoge, 1973), the degree of
elasticity, secant modulus (a/c at 3% compression), toughness
(work per unit volume necessary to compress the sample to
25% of its original
thickness),
and work ratio
(ratio of
the area under the load curve of the second cycle to the
area under the load curve of the first cycle) are much
lower than the values of these same rheological properties
obtained at other water activities.
These results indicate
that a greater degree of compresion of precooked beef
can be expected at these levels of moisture content.
Kapsalis et al. (1970) also indicate that as the B.E.T.
monolayer value for water sorption is approached, rigidity
(secant modulus), toughness, bioyield strength, work ratio,
and degree of elasticity reach a maximum which explains why
at very low levels of moisture, foods crumble upon compression.
It should be noted that the effect of water at different adsorption levels is strongly dependent on temperature and basic material structure.
Many polymers undergo
what is termed a second-order phase transition as their
temperature is raised.
The change has been described as
an internal melting in which the polymer changes from a
brittle
(glassy) to a rubber-like
state, and many of its
physical constants alter in value.
Presence of moisture
(or other low-moelcular-weight solvents) lowers this
transition temperature.
It is thus possible that increasing
39
concentration of these agents may even cause such structural
change when the sample is held at constant temperature
(G. King, 1950).
At this transition temperature most materials
start to flow and lose their original solid structure.
Thus, increasing the sample temperature is also able to
increase the mobility of the polymeric segments.
Sakata and
Senjue (1975) reported on thermal compressibility of thiolignin and diaxone lignin in the presence of various lowmolecular-weight solvents including moisture.
Powdered
material containing a predetermined level of the solvent
was compressed to form a pellet which was then placed in
a flow tester.
When the temeprature was elevated at a
constant rate of 3C/min
under a constant compressive load
of 10 kg/cm2, they detected three regions of compressive
deformation.
Noted first was a softening region in which
small voids are reduced as the sample gradually deforms
under the load.
At the softening temperature (Ts) , the
small voids are eliminated and the pellet becomes a homogenous, transparent material or a single solid phase.
At
temperatures just above Ts there is little additional compression taking place and no apparent change in the pellet.
At some higher temperature (Tf) the flow region is reached
and the sample starts to flow and passes through the
nozzle which is located in the bottom of the tester.
It should be noted that moisture and temeprature
influence the rheological properties of polymers differently
40
depending on the basic material structure.
Basic material
structure relates to the composition and organization of the
components in food.
In particular it is concerned with
structural polymers (such as cellulose in plants and
myofibrils in animal tissue), and those relatively lowmolecular-weight compounds that are located between the
polymeric segments and have a low glass transition or
flow temperatures (such as sugars).
For example, sugars
seem to be critical in affecting the mechanical properties
of plant materials, since they are mostly amorphous after
freeze drying and can absorb more water in the amorphous
state than the cellulosic materials.
At
sufficiently
high water activities or temperatures, they will soften and
flow under a compressive force.
If they flow too readily,
the cellulosic phase may deform to such an extent that it
may lose its ability to retain a reversible strain
(MacKenzie and Luyet, 1971).
This will cause collapse of
the material under the compressive load.
2.6
Methods of Compression and Debulking
Dehydrated Foodstuffs
To accomplish compression of foods without shatter-
ing and fragmentation of the dried matrix, it is necessary
to have the solid structure of the material in a "compressible
state."
As mentioned earlier (sec. 2.5), depending on the
basic material structure, this state can be achieved by
41
manipulation of combinations of moisture content, temperature,
and presence of other agents.
In the compressible state the relative rigidity of
the polymeric segments within the cellular structure of
the food is sufficiently low to allow deformation of the
solid matrix without irreversible structural changes under
a compressive
load.
This state can be achieved by bringing the dry food
to a relatively uniform moisture content in the range of
5-20 percent depending
on the food product
(Hamdy, 1960).
Moisture may be transferred either as vapor or sprayed as
a liquid mist.
It has been shown that the amount of moisture
added to the dried product prior to compression significantly
affects the rehydrated product (Rahman et al., 1971).
For example, as the conditioning moisture increases,the
rehydration ratio decreases whereas the resistance to shear
becomes greater.
For some products manipulation of temperature alone
can be used to achieve a compressible state (Rahman et al.,
1970; Do et al., 1975).
In some early investigations
there have also been attempts to combine the effects of both
moisture and temperature to accomplish compression (Donnelly,
1947; U.S.D.A., 1948).
Problems associated with thermal
treatment are discoloration, burning, and loss of nutritional
value of the product.
Agents other than moisture,such as as ethylene,
propylene glycol, glycerol, vegetable oil, etc., have been
42
used with some success for compressing different food
products (Gooding et al., 1958; Durst, 1967; Ranadive et
al., 1974; Pavey, 1975; and others).
The major limitation
of using these agents is the foreign flavor that they
impart to the food.
Once the materials are treated (by humidification,
heat, or other agents), they are then placed in a compression
cell in a hydraulic press and subjected to uniaxial pressures
ranging from 100-5000 psi.
The samples are then redried
or cooled and packed to yield the compressed product
(Rahman et al., 1969).
Freeze-dried foods high in quality, nutritional
value, and recovery characteristics can be obtained using
the currently practiced methods for producing compressed
foods.
The method has been described by Rahman et al.
(.1969) and is shown diagrammatically
in figure 2-3.
2.7
Alternative Approaches for Production of
Freeze-Dried/Compressed Foods
2.7.1
Limited Freeze Drying
Most developmental work to date for freeze-dried,
compressed food has utilized freeze drying to low moisture
contents,followed by remoistening of the dry material to
attain the desired moisture content for compression.
Under
certain controlled conditions it is possible to conduct
the freeze dryingprocess such that at the "end" of the
drying, the material has a uniformly distributed, desired
43
o
Z
LUw
or
-O0
z
0
4
o
U,
m
CD
CD
z
zC
w
U
Z
)- ->
o
-J
r
0o
>z
4W
Ir
Z
4
t
CD
W
C
rd
a)
U
,q9
z
o
I
Z
rl
0 ck
ow
CD
,4
0.
I
C
LiiC
-tz
L
I
0
N
b.
W
t
< W
-W
- N
N
a)
z
4-
LL
Z
CW
iL
toz
I
-i
O
0
z0N
'UJ4
IL
(/-
rC
H
r
O
Q)
rd4
CI)
0Ili
Xi r
-P
L
(U
rnm
rd0
Z0
-N
)'W
z
l
a
.H
c-
uw
bW
44
moisture content, thereby avoiding the remoistening step.
The term "limited freeze drying" was first applied by
MacKenzie to this process (MacKenzie and Luyet, 1969).
The principle of the method is to control the freeze drying
such that the relative humidity at the surface and within
the product does not fall below some predetermined value,
which is high enough to ensure that throughout the drying
the equilibrium moisture content in all regions of the
product will not drop below the desired value.
A major limita-
tion in this method is that the sample temperature and the
condensor temperature must be controlled within very narrow
limits (+ 0.1°C).
temperatures
Generally, the selection of sample chamber
in the range of -10 to -40°C permits
tion of materials
containing
the produc-
about 25 to 5 g H 2 0/100 g
solids, respectively (MacKenzie and Luyet, 1971).
To
accomplish limited freeze drying through control of environmental relative humidity requires that the temperature
difference and water vapor partial pressure differences
(driving forces for heat and mass transfer) be less than
in ordinary commercial freeze driers, which necessarily
leads to longer drying times.
King et al. (1975) noted
that MacKenzie and Luyet had reported that drying times of
25-100 hours were required for 1 cm cubes of beef and various
vegetable products.
King et al. (1975) reported that drying rates for
limited freeze drying processes can be accelerated if:
45
*
The heat transfer coefficient from the heat
source to the surface of the product is increased.
*
The temperature of the heat source
during the early part of drying so
temperature of the product surface
the point where the surface itself
the prescribed relative humidity.
*
Control of the environmental relative humidity
is accomplished through other means than product
and condenser temperatures.
is raised
that the
rises toward
experiences
They proposed two alternative methods to accomplish
a more efficient limited freeze drying process.
In one
method the heating platen temperature is controlled at
the desired final surface temperature, instead of the sample
chamber temperature which was originally used by MacKenzie
and Luyet
(1969, 1971).
They demonstrated
a 50% reduction
in time required for limited freeze drying when compared to
a constant platen temperature drying process.
A major limita-
tion of this technique is that small fluctuations of the
platen temperature will result in a significant change
in final moisture content through its influence on the
local relative
humidity
(King et al., 1975).
The second method described by King et al. (1975)
is a self-regulating control of the enviornmental relative
humidity by using a water-uptake medium which maintains a
particular, predetermined relative humidity independent
of temperature or amount of moisture taken up.
At the end
of the drying process, the sample surface and water-uptake
medium relative humidity would be the same, provided that
food pieces and water-uptake medium are at the same
46
temperature.
Jones and King (1975) described a system
with alternating layers of food and hydrating salt desiccant.
A circulating low-pressure gas is used to transfer heat and
mass between the food and desiccant.
Since the latent
heat of absorption for most desiccants is about equal to
the latent heat of sublimation for the same amount of
water, the gas will be cooled as it passes through the
food layer and then will be dried and reheated as it passes
through the desiccant layer.
In the process of limited freeze drying when the
food is removed from the freeze drier, it is ready for
immediate compression and subsequent final drying to low
moistures as shown diagrammatically in figure 2-3.
2.7.2
Microwave Technique
In this approach compression of food products could
also be achieved without the need for the initial overdrying
and humidification steps which are required by the conventional processing techniques.
This technique is based
on stopping the freeze drying process at a time when the
amount of ice remaining in the food is approximately 20%-25%
by weight of the total solids.
The condition of uniformly
distributed moisture which is required for compression is
obtained by vaporizing the remaining ice by a microwave
treatment.
Haralampu (1977) reported that partial drying
of peas to 20%-25% moisture could be achieved, reproducibly,
at a platen temperature of 1620 F (72°C) and a drying time
47
of 12.5 hours.
Microwave treatments at low power levels,
but high energy inputs,were sufficient to vaporize the
remaining ice and redistribute it relatively uniformly
throughout the sample so that compression could be achieved
without damage to the solid matrix.
He noted that the
microwave processed product compared well with conventionally
compressed product, with regard to rehydration ratio, fines
production and volume recovery.
Besides eliminating the
overdrying and humidification steps, this process also
eliminates the slowest and most costly period of the
freeze drying process--the desorption phase when all the
ice has been sublimated.
Once the microwave treated samples
are compressed, they are further dried to produce the
dried/compressed product (figure 2-3).
Problems associated
with this technique include reproducibility of achieving
the desired final moisture content for the partial freeze
drying, and discoloration, and burning that can occur during
the microwave treatmentwhen higher power levels are used.
2.7.3
Simultaneous Freeze Drying and Compression
The phenomenon of simultaneous freeze dehydration
and compression was first reported by Konigsbacher (1974).
In this approach the food material is subjected to compressive force during the freeze drying process such that
drying and compression of the food material take place
simultaneously.
This avoids many processing steps such as
overdrying, rehumidification, redrying and other treatments
48
that are required in the other techniques used to produce
freeze-dried/compressed foods (figure 2-3).
A wider
demonstration of the feasibility of this technique has been
made by Emami (1976). He noted that his results indicated
that during freeze drying there exists enough moisture in
the diffusion zone at the ice-dry layer interface to make it
suitable for compression.
Application of force thus results
in a gradual compression of the interface as the drying of
the food material proceeds to low moisture contents.
A spring-driven device was used to provide the required compressive pressures in the range of 5-185 psi
(figure 2-4).
The results found by Emami et al.
(1979) are
summarized in the following sections (2.7.3.1 to 2.7.3.6).
2.7.3.1 Compression Behavior
For a variety of food materials tested at fixed
sample thicknesses, the compression ratio was linearly related
to compressive pressure over the range of pressures tested.
However, the actual compression ratio obtained at a given
force was dependent on the particular food item.
Figure 2-5
illustrates the relationship of compression ratio and compressive pressure obtained for green beans.
Theoretical
considerations would indicate that for a reversible compression (i.e., without destroying the solid structure), the
compression ratio will reach a maximum value at some
compressive pressure so that no increase in compression ratio
49
Fig. 2-4.
Schematic of spring compression system (loaded).
I
Base
Plate
2
Top
Plate
3
Matched Set of Springs (Extension Type)
4
Threaded Guide Rods
5
Nuts for Raising Top Plate
6
7
Sample Cell Stand
Sample Holder
8
Compression Plate
50
o)
a)
a)
tn
0
U)
-P
cd
-H
0
U)
a)
,CL
a-
o
U)
4U)
0-
04
O-H
O
v In
U
(4 44l
FZ4
011.V
NOISS3:dWOO
51
occurs with further increase in compressivepressure.
It
is thought that this maximum compression ratio, which
reflects the "true" volume change within the given sample
piece and not the voids between individual pieces, is directly
related to the-initial water content of the food material.
As noted earlier, this is measured by the "displacement"
compression ratio.
2.7.3.2 Effect of Sample Thickness
When the initial bed thickness was varied for green
beans, a plot of compression ratio versus sample thickness
resulted in two linear, parallel lines corresponding to the
two fixed compressive pressures tested.
As shown in
figure 2-6, compression ratio decreased linearly with increasing sample thickness for each given compressive pressure.
Since the slopes are equal, this decrease in compression
ratio is apparently independent of compressive pressure,
and apparently is due to geometry effects for a bed of
green beans when packed in the holder.
In packing cylindrical
bodies,as the number of layers are increased, the ratio of
end layer void volume to the total void volume will not
stay constant.
This results in increased packing density of
the sample, since as the number of layers increases, the
void volumes between the pieces are a smaller percentage of
the total volume.
A similar view is obtained by noting
that when packing cylindrical bodies such as green beans
in a rigid holder, a lower packing density of material at
52
0 10
40 psi
on
96 psi
0
O
8
A~~
-
6
A~~
\.A
w
V)
zn
1%,
4
0
0
w
0
2
0
0
I
I
I
I
I
I
2
3
4
5
NUMBER OF LAYERS
Fig. 2-6.
Dependence of compression ratio on sample
thickness for green beans.
53
the walls of the sample holder, or in this case the top
and bottom plates, will result.
As "internal" layers are
added, the effective initial packing density of the entire
bed increases.
2.7.3.3 Final Moisture Contents
Emami (1976) showed that the final'moisture contents
of compressed samples were always similar to those of noncompressed controls, freeze-dried at the same time.
In most
cases the final moisture contents were in the range of
1.5-5 percent, which indicated that compression during
freeze drying did not prevent dehydration to low moisture
levels.
It was further noted that final moisture contents
did not seem to depend on the compression ratio.
2.7.3.4 Rehydration Behavior
Emami (1976) also investigated the rehydration of
compressed samples and showed it was quick, in most cases
reaching equilibrium within 15 minutes.
Samples generally
reattained their precompression volume after rehydration.
It was shown that when just immersed in water, freeze-dried,
non-compressed vegetables generally rehydrated to a lesser
extent than compressed samples.
However, evacuation of the
freeze-dried samples prior to addition of water showed
little effect for compressed samples, whereas the noncompressed samples showed a marked increase in rehydration
(figure 2-7).
Vacuum rehydration increased the extent of
54
O
.
0
0)
ro
n
I
])
0,
N
0)
0)
U)
o-
4o
0
*rl
-P
.S.
0)
CC
0
to
-P
E
w
2_
rd
-,-
to
rd
0
a)r.
CN
I
(,
(q4sa
) S6
OH
oZH -5-/
/(Ap)
S1 6
OZH-6
55
rehydration of non-compressed vegetables to that of compressed
samples.
These results indicate that the lower uptake of
water by non-compressed vegetables was due to air entrapped
in the cells and that some structural changes which are
caused by compression affect either removal of the entrapped
air in the cells or water uptake by the cells resulting in
improved rehydration behavior.
It was also.noted that for vegetables, the extent
of rehydration was not affected by the degree of compression
(figure 2-8), whereas beef cubes showed a slight trend
toward reduced rehydration with increasing compression
ratio
(figure 2-9).
2.7.3.5 Organoleptic Evaluation
Organoleptic evaluations of cooked,non-compressed,
compressed, and fresh or frozen samples indicated that
compressed samples were always rated equal to or better than
the respective non-compressed samples.
Fresh samples were
usually rated higher than either of the freeze-dried products,
though in most cases the differences were not significant
at a 1% level.
2.8
Moisture-Temperature Gradients in the Dry Layer
during Freeze-Drying; Their Effects on Mechanical
Properties of the Dry Layer during Freeze Drying
State of water during freeze drying has already been dis-
cussed.-(section 2.3).
The idealized distribution of moisture
during sublimation is that shown in figure 2-10a (Karel, 1975).
56
__
-
-
m
I
.~~~~~~~~~~~~~~~~~~l
lo
I
ro
I
0
I
4-1
I
I
,4
U2
0Q
Q1
O
0
0
.rr
_
II
l,
0
l
3
U)
w0
co
I
°l
-I
0
U
f
I
I
U)
4-
0
I
Q
00
-I1rd
I
o
0
I
c4
'C
.o
c
(q(S9)S
,
.
*
g-P4
-4
r
-o .
I
a0
I
o
OZH 6 /
i
o/(
a
0 0
i
o
) Sl
i
I
ci
'6
O0ZH'6
0
.I
o
0o
57
a)
a.
N
/
/
4-I4
0
0
0
-I1
/
no
4
4,0
/
0 9.
O
V)
0
J
A
N
U)
/
U)
-:
z
0
/
O)
4.
O.
0
a,
440
G
a0 G
a)
0
Q.
I..
4)
/
0a
4l-
I
E
-O
!
t
0
0
4A
I
I
LIu4
I
0s
-
0
o
o0
I
I
dd 0
II
0 0o
d
(qse
OH
O~H
./'S' OH('qJ)'D
O~H '6
'"
u
58
in which there is an interface at which. the moisture content
drops from the initial level (m0) in the frozen layer to
the final moisture content of the dry layer (mf).
The final
moisture content (mf) is determined by the equilibrium of
the sample solids with water at the partial vapor pressure
of water in the space surrounding the dry layer.
However,
evidence given by many authors indicates that this is
actually not the case,and the moisture profiles are represented more accurately by figure 2-10b, which shows the
existence of some gradient in the "dry layer" (Karel, 1975).
Temperature histories during freeze drying of
different materials have been reported by many investigators
(Kessler, 1962;
Sandall et al., 1967; King, 1971; Aguillera
and Flink, 1974; and others).
Their data indicate a very
thin transition between a frozen zone of uniform low
temperature, and a dry zone with a nearly linear gradient
of temperature between the ice front and the outer surface.
Gentzler and Schmidt (1973) reported a linear relationship between log of the local moisture content at any position in the non-ice portion (referred to by the authors as
the bound water at that position) and the temperature difference between that position in the dry layer and the.temperature of the frozen region during freeze drying of skim milk.
Their relationship was based on higher desorption temperatures for bound water as compared with the sublimation
temperature for frozen water.
Aguilera and Flink (1974)
59
H
z
Lw
d
0
U
en
;5
tO
0)
2
(a) IDEALIZED GRADIENT
I-
Z
Ll
o
2h
(b) PROBABLE
Fig. 2-10.
GRADIENT
Models of moisture gradients present during
freeze drying (from.Karel, 1975).
60
showeddata of Oetjen (1973) for coffee grounds gave similar
behavior.
This behavior is shown in figure 2-11, for both
the skim milk and granulated coffee studies.
From figure 2-11 it can be seen that if the temperature of the frozen region remains constant during the freeze
drying process, then at any instant in time, the local
temperature and moisture content of any location in the dry
layer zone are directly related, and thus for each given
temperature in the dry layer there will be a unique value
of moisture content at that location whose value can be
determined from figure 2-11 (for skim milk and granulated
coffee).
Aguilera and Flink (1974) report that for freeze
drying with constant surface temperature, a "rotation" of
the temperature gradient occurs about the constant surface
temperature point, resulting in decreasing temperature
gradient as drying proceeds.
Since the moisture content
remaining after the passage of the ice front is directly
related to the difference between local temperature and the
ice front temperature (as described above), higher moisture
contents can be expected at a given distance from the ice
front as drying proceeds.
This is in accordance with. descrip-
tions of Bralsford (1967) and Brajnikov et al. (1969) of a
soft wet broadening zone between the ice zone and the dried
layer, since intermediate levels of moisture build up
continuously against the front as drying proceeds.
61
N
z
00
o
P
z
I
r
i
(K)
dflerence
Temperoture
Fig. 2-11
Local moisture contents in the dry layer
during freeze-drying as a function of difference
between local temperature and ice-front temperature (from Aguillera and Flink, 1974)
62
The effects of moisture and temperature on mechanical
properties of foods and other polymeric compounds were described
previously (section 2.5).
By combining these effects with
moisture and temperature gradients and their relationship
in the dry layer and at the ice front, it can be expected
that at any instant in time, during freeze drying, each
location will have different and fixed mechanical properties
which is a direct function of the moisture content and
temperature at that location.
Furthermore, since moisture
content and temperature of each location in the dry layer
will change as the drying proceeds, the mechanical properties
of that location will constantly change with time during
the freeze drying process.
It can be expected that at
any instant in time,if a constant compressive load is applied
to the food during freeze drying, each location in the dry
layer will undergo a different degree of compression.
These considerations which have not been investigated previously
are an important aspect to the compression of foods during
freeze drying and are investigated in this thesis.
63
3.
EXPERIMENTAL
The experimental procedures have been divided into
twomajor parts.
The first part presents procedures for
experiments involving compresion during freeze drying,
while the second part presents the materials and methods used
in evaluation and characterization of the compressive
theological properties of product as a function of moisture
and temperature (referred to as "Instron studies").
3.1
Equipment for Compression during
Freeze Drying
3.1.1
Hydraulic Compression System
The hydraulic compression system was designed for
complete insertion within the freeze dryer chamber.
A
condensor system was developed which allowed collection of
the sublimed ice so that drying rates could be determined,
since direct weighing was not possible due to the large
mass of the hydraulic cylinder and cell support assembly.
A schematic of the hydraulic compression system is given
in figure 3-1.
The hydraulic compression system operates by means
of an air driven pump (SC Pump Model 10-500-3), which develops
hydraulic pressures from 0-2500 psi from laboratory air
up to 45 psi.
Control of the air pressure regulator determines the
fluid force (up to 5000 lbs) applied on the piston of the
64
Fig. 3-1.
Schematic
of the hydraulic
compression
system.
IO
0
0
1
10.
1.
Chamber Top (plexiglass)
2.
Chamber
11.
3.
4.
5.
6.
7.
8.
9.
Hydraulic Cylinder
Sample Holder
Chamber Base
Hydraulic Pump
Hydraulic Fluid Lines
Valve (bypass)
Vacuum Isolation Valve
12.
13.
14.
15.
16.
17.
18.
Thermocouple Wires
Heater Power
Vacuum Gauge
Valve
Valve
Drierite Tube
To Vacuum Pump
Condensate Collector
Chilled Alcohol Bath
65
hydraulic cylinder (Hennels Co., Model HH-144-N-2X1).
Thus
while the maximum compression pressure in the hydraulic
cylinder is fixed, the ultimate compression pressure available
at the sample will be related to the sample holder surface
area.
(If required, this system can be easily converted
to higher forces or more sensitivity at lower forces by
changes in air pumps or cylinders.)
To distribute the
hydraulic force, the cylinder is fixed to a top plate-base
frame and the cylinder rod acts on the "compression
plate" of the sample holder.
3.1.2
Vapor Trapping System
Figure 3-1 also shows the schematic for the vapor
trapping system which permits determination of drying rates
during compression/freeze drying.
The vapors sublimated from
the sample are condensed in a stainless steel vessel immersed in a chilled alcohol bath.
The stainless steel
vessel is removed from the trapping system by isolating the
chamber (which remains under vacuum) from the condenser
and bleeding in dry air (close valve.9, 14; open valve, 8, 13).
A transfer of condenser vessels can be achieved in about
30 seconds with very little increase in chamber pressure.
A materialbalance showed in almost all cases that the collected condensate is approximately equal to the material
sublimated.
66
3.1.3
Sample Cell Holder
Sample holders have been constructed of mesh or
perforated metal sheets having different perforation sizes,
and percentage open areas to facilitate vapor transport
through the walls as well as the top and bottom surfaces.
A schematic of the sample cell holder is shown in figure 3-2.
Parts A and C are termed top and bottom compression plates,
respectively.
The fine wire thermocouples connected to these
plates are used to both monitor temperature of the plates
and as inputs to the heater controllers which are controlling
the amount of heat input to the sample being freeze-dried/
compressed.
In most experiments the percentage open area
of these plates was chosen to be 63% (maximum available)
to maximize vapor transport through the top and bottom
surfaces.
The top and bottom compression plates rest on coarse
metal gratings which provide sturctural support for the system
when it is under hydraulic force, while not hindering
moisture transport.
The cell wall (Part B) is constructed either from
the same perforated metal sheets as the compression plates
(to obtain moisture transport from the sides) or from solid
material (no opening) in order to obtain a more uniform
drying with moisture loss from the top and bottom surfaces
only (i.e., closer to the U.R.I.F. model).
67
Top Compression
Plate
(A)
Cell Walls
w
(B)
Thermocouples
i
ol/
I
Bottom
Plate
Fig. 3-2.
Compression
(C)
Schematic of the sample cell holder for
compression during freeze drying experiments.
68
The sample cell holders were constructed in three
different sizes (7.5 x 7.5 cm 2 , 10.5 x 10.5 cm
,
and
15 x 15 cm2 ) in order to evaluate the effect of size on compression and freeze drying of foods.
In operation, the bottom compression plate (C)
rests on a metal grating which is stationary while the
top compression plate (A) moves downwards inside the cell
walls (B) as the sample is being compressed.
The location
of the top compression plate (and thus the sample thickness)
is monitored continuously by measuring the voltage drop
across a linear variable resistor which is connected between
the hydraulic cylinder support and top compression plate.
When controlled heat input is desired, silicon
rubber heaters are placed above the top and below the bottom
metal support gratings mentioned earlier.
Electrical energy
to the heaters is controlled by thermocouples located
either on the heaters or at chosen locations in the sample.
Thus, for this system, the heat transfer is both by conduction
(through the metal gratings and compression plates), and
radiation (through open spaces of the metal gratings and
perforated metal sheets).
3.1.4
Sample Loading
Cut, frozen green beans purchased at a local super-
market were hand sorted to obtain regular cylindrical
pieces of adequate dimensions.
The green bean pieces were
69
stacked in the sample holder in three layers like a pile
of logs so that the compression force was only in the radial
direction.
3.1.5
Program for Compression during
Freeze Drying Experiments
The experimental program is designed in such a way
that the effects of parameters relevant to compression and
drying behavior of the product could be investigated independently.
The experiments conducted in this phase of
study are listed in table 3-1.
As can be seen from table 3-1, parameters of interest
are compressive pressure, heat input, and size and porosity
of the sample cell.
3.1.5.1 Compressive Pressure or Loading
Compressive pressure denoted in psi indicates the
amount of pressure load delivered by the hydraulic compressor
on the sample, this pressure being held constant at a given
value throughout the experiment.
In order to evaluate
the dependency of the degree of compression on sample
loading or the applied compressive pressure, different
levels of applied compressive pressure were used in different sets of experiments as shown in table 3-1.
Further-
more, since compression of the dry layer during freeze
drying is expected to result in improved heat transfer
70
m
03
CY)
o
m
m
m
C
(
ZQH
44
0
H
k
o
Ln
Ln
r.
LA
Ln
L)
*.
Ln
Ln
Ln
Ln
o
o
X
X
4-i
0
H
H
.
-i
U2
En
i
r-4rU N
) 0-H O
U-)
LA
"1
i-
t-
-H
N
.
I,
LA
"
LA
.*
LA
.
LA
.
LA
.
*
O
N
N
N
N-
N
N
H
0
U)
i
04
N
o
>1
IN
04 -U)
0-H 0 0U)
O
4N) 04
dP
riP
LA
d
o
cm
dP
c
do
m
dP
CY
dP
m
dP
m
rP
tD
C,
N
ye
N
a4
cLl
O
4-4
U)
04
-H
q 4
0
04
Ul
U1)
Ull
U)
N
U1)
)k
Ad
O
o
(
r.
~4
a)
U)
U)
U)
0
tn
r
O
r.l
C-)
0
r4
La
a)
N
P4 4-
0k0rp
r-
U
3
-
kD
m'
e
'0
'D
m
c
o
o
mo o
U)
O.
C,
0
U)
k
O
4-J
-IN
U)
l U)
U)U
0 O
uH14t))
0
H
o
M4-)
U
o
sr
o
F:0
) O-
O 044-'-
N U)
-N
p4
0U)
U)
En
U
r-i
0
.H
a)
U
o
s
c
U
Io
O
U
O
O
H
O
H
U
O
H
U
O
H
U
O
H
O
O
O
O
-.
o
U)
En0
AI
a)A
U)
{o
U) -
o
E-1
O0
U
U}
)
I --
m
N U)-.
O
o
C
O'
o
o
or-
U)a
rt U4
o
k4-Nb
U.
0 NO
-H
aH
aU
U) 4#
)
MN
O
LA)
om
O
o
N
V
I
LA
L
Ln
CN
d
LA
LA
#0U-U)
4-)
U2
f V]
H
c
C)
Lr
-H
Q
N
04
U)
U)4)
NU)
71
(due to increased thermal conductivity of the dried layer
and reduced pathway length for heat transfer), this effect
must also be considered.
3.1.5.2 Heater Controls/Energy Input
Heater controls relate to the amount of energy being
supplied into the system in order to accelerate the process.
The temperatures indicated in table 3-1 correspond to
set points for the controlling thermocouples located at
the inner side of the top and bottom compression plates (in
touch with sample surfaces) as shown in figure 3-2.
Evaluation
of the influence of this process variable provides information on maximum or optimum heating plate temperatures
and/or surface temperatures to accelerate drying rates,
achieve high degrees of compression and yet not exceed the
limits for a given set of mass transfer and pressure load
conditions at which irreversible structural changes begin
to occur.
3.1.5.3 Sample Holder Size and Porosity
Size refers to the length and width of the sample
holder used in this study.
These dimensions were chosen
such that the surface areas have a ratio of 1:2:4 (i.e.,
2
2
2
7.5 x 7.5 cm 2 , 10.5 x 10.5 cm , and 15 x 15 cm ), which
also means that the weight loading in these holders has
the same ratio.
72
Porosity is the percentage open area of the sample
holder cell walls (figure 3-2).
As mentioned earlier the
percentage open area of the top and bottom compression
plates was fixed at 63% which is the maximum porosity
available for metal sheet of sufficient strength.
Porosity
of the cell walls were varied as shown in table 3-1.
These variables (size and porosity) relate to
potential mass transport behavior, both in percentage open
surface area and relative diffusion distance in the x-y vs.
z planes.
To overcome the effect of compression on the
potential for vapor buildup in the dried/compressed layer
and reduce the effect of compression plate on the drying
surface, sample holders and compression plates of special
design are required so that the material can be subjected
to surface forces without extensively impeding the flow of
vapor from the drying surface.
3.1.6
Terminology for Compression during
Freeze-Drying Experiments
Henceforth, parameters for a given experiment will
be denoted in the following manner:
dimension,
% open area,
0 C,
psi
Forexample, 7.5, 0%, 10°C, 40 psi would mean an experimental
setup where the size of the sample holder is 7.5 x 7.5 cm ,
the side walls are solid (0% open area), the controlling
73
thermocouples on the top and bottom compression plates are
held at 100 C, and the amount of compressive pressure is
40 psi held constant throughout the experiment.
3.2
Materials and Methods Used in Evaluation
of the Rheological and Mechanical Properties
of Greenbeans as a Function of Moisture and
Temperature (Instron Studies)
3.2.1
Freeze Drying
Cut, frozen green beans purchased at a local super-
market were hand sorted to obtain regular cylindrical
pieces of adequate dimensions as before.
A conventional
laboratory freeze drier (Virtus, Uni-Trap Model 10-K) was
used to freeze dry unilayer batches of green beans to
low moistures at room temperature and a chamber pressure
of below 50 millitorr.
After freeze drying,the green
beans were kept under vacuum in a desiccator over Drierite
before rehumidification.
3.2.2
Rehumidification
Seven saturated salt solutions of known water activity
(aw ) (table 3-2) were placed in the bottom of seven desiccators.
The freeze-dried green beans were then placed
above the salt solutions in mesh containers.
The desic-
cators were evacuated to hasten the equilibration process.
The desiccators remained covered at room temperature (20QC)
for about one week.
74
Relative humidites used for rehumidification of freeze-dried green beans for
studies of mechanical properties
Table 3-2
Salt
Drierite (not in solution)
%RH
0
MgC1
33
NaBr2
58
CuC1
66
NaCl
75
(NH4 )2S04
81
KNO3
93
Source:
Labuza et al, 1976
75
3.2.3
Compression with the Instron Universal
Testing Machine
A compression cell made of plexiglas (figure 3-3)
with dimensions of 2 cm x 3 cm was used for compression
tests.
A compression block, also made from plexiglas, was
constructed to exactly fit in the compression cell.
The
cell was designed so that three green bean pieces having
approximate dimensions of 2 cm x 3 cm would make one
layer (figure 3-3).
Three layers of green beans were used
for each test (i.e., total of 9 pieces).
The rehumidified samples of green beans at different
relative humidities were packed in the compression cell as
shown in figure 3-3 and quickly adjusted to the desired test
temperatures (-250 C, -10 0 C, +150 C, and +240 C) before compression.
Eight to ten samples (9 beans each) were compressed
at each specific combination of moisture content (RH) and
temperature.
An Instron Universal Testing Machine (Model 1122),
equipped with a 500 kg tensile-compressive load cell,was
used for compression tests.
The applied force was recorded
as a function of the crosshead position with full scale
readings
of 10, 20, 50, 100, 200, and 500 kg.
The three layers of green beans were compressed
uniaxially to a final thickness of 3 mm at a crosshead
speed of 20 mm/min.
To evaluate the effect of crosshead
speed, in another set of experiments, freeze-dried green
76
)RESS IO'N
K
AIR VENT
All
VE
Fig. 3-3.
Schematic of the sample cell holder for compression tests using the Instron Universal
Testing Machine.
77
beans rehumidifed at different relative humidites were compressed at room temperature (24C)
of 50 mm/min.
Once compressed, the stress-strain behavior
duringunloadingwas
time).
at a crosshead speed
immediately measured (i.e., no dwell
The crosshead speed in the reverse direction
was the same as the downward direction.
After the compression, the sample was dried in a
vacuum oven at 65C
(1500 F) for 24 hours for determination
of the moisture content.
78
4.
RESULTS AND DISCUSSION
Analysis of the experimental results have also been
divided into different parts to follow the organization of
the experimental procedures.
The major categories include:
results of experiments involved in compression of green
beans during freeze drying, evaluation of the compressive
properties of this product as a function of moisture content
and temperature (Instron studies), and finally correlation
of the behavior of compression during freeze drying to the
mechanical properties of this product.
4.1
Compression during Freeze Drying Results
Table 4-1 shows the overall results obtained in
different sets of experiments conducted at different values
of the parameters mentioned in section 3.1.5.
4.1.1
Dependence of the Degree of Compression on
Sample Loading; Compressive Pressure-Strain
Relationship
The dependence of the degree of compression expressed
as strain ( = AL/L) on the amount of compressive pressure
applied (compressive pressure-strain relation) during
freeze drying is illustrated in figure 4-1 for different
sets of experiments.
It can be seen that at low levels of
compressive pressure (i.e., less than 10 psi), very little
or no compression can be expected to occur for green beans,
and a minimal amount of compressive load is required to
79
0
U)
0
m
r- 0
N
C4 CS m
C
4
N4
N
LON
*H 0
000
0000
0 00 0
NI
a)
U
0
0000
l
rl C TN
0000
oo o o
01 - 0 oa
a
.)
0
rlH
U
N
U
t~)
X
a)
N
a) * U
0-i 0H
u)L)
Ln L
x
0000
000
oLn o
:
0
)
o
0
O
o ~3IVoo
or osr O Oo
Ž0
.O-I)
U~
· ) 4-
00N Lfl
xxx
XXXX
LA LA LA LO
LuLAL
LOLOLC)Ln
N
N NN-
LLOr)
LA Ln Ll
d
U) -
rP > r
.14
rd
dP
Ln
m1
*HUla,
a)
En (
00
dP
Od
O dt°
0
IO
I
U)N
)W
$4
a)
40
PdPdP dP
ln
Lu Ln
mc
Ln
m m
.1-
a
E-4
00
a)
re)
dPd
en m
r1,
9
9
dP dP dP dP
m () cn c
%.0 I\o D
9L
)'IO
dP d
dP
m m m CO
~0 'I %0 0
dP dP dP
~ m m
D k0 'I
,
a)
N
0
U
U
0 00
UUUU 0
000
a)0
0
0c .) _HQ) rl
dP
dP dP dP dP
03 0 0 CY
: 'O %0
k
dP dP dPdP
0 0
0
kO O kO
0
1-4
a)
O
dPdPdP
cr)
110 I'
0000
IDc
+ (?
+
C
I eI eI
(?
UUUU
0000
+
0
0
0
0
0
0 O tQ
.H
U)
a)0
N
-)-U)
l
En )
a) t
O O 00
P4 0
41:
4.)
a)
(n
c, un
41: 1 1:
41: 4 #:
CN
00 0 0
4.
i P4
04
P404
~ ~~t~~
114:
4
L00
0
41:
) ~~
c
4=# --
4i
4P
~t
0000
000
o
3g r
(Nd~~DN
ml~ou
41:41:41:
cr
~~~41:
)-
~
Ux0
)
cW ~~tW
1:
LA 41: 41: *4#--
41:
. 4.)
Ux0
) 9~
El)
0 0
4
t~~t
80
O
2r
u
o\
.H
0
N- I
4m
aJ
0 m "T N L A v Nm(N
(N )M, (4
C
C
~
4J
C-1
4)II
a 2W
z 4
a
a
Q
a
L N N N N Ln
O
H o 'O
NNN-r-Ln
0000 C C
0
U}
>0
r4
I
~l ~q
t q~
I
l qlrlu)C~~mS
O q C m
q
O
co
4)
10
-
a)
( 3
Cd
--
h
a
0
0
.H
4-) a
0
0)
O
O
0 00 00 0 Cf 0 0 0
M
,-I.,. r
0
H 0H 0U
u
r0
V
a)
H.
H
4-I
a
o
4)
O
U
N
a
.C
rd
o0 L
a)
N
-H
·H
0
x
o
0
o
Lo
.
O
{
4k-
Cd
LA
Ln
.0
·o
UA -
L
LA
.0 @o.o
.0 .0
.lJ
9
rlU) Oz
fttt
dP dP
00
NCIC
OQ
0
dP d dP dP dP
-r m KD xD -10 'I
A
a
OdP
C4
.0
m
.H1
C.o
O
'I tD
dP
dP
t9 t9o 'I 'I %
dPdPd dPdPd dPdP
000000000
_
.4
a)
H
-a
>
O
a
3
0
4-)
0
o
O
m
U0
O
i--
a
0
A
a)
N
-.)
4)
02
I-q
U
uo
.4)
kV
aI)
H
w (rl
0202 -
UUUUUUU
-H
0
H
Cd
4)
U
o o o o o o o U Uo U
00000000
o
00 o
aEw
-01 ,00i
- i-
-H
oz
.1i
0 0 N0
V
02
0 04
2 a)
a)o*
0
a,0
:
U1
>
-H
00
4ro
a)
4-)
C1)
a)
44
000000o0000
0a
..
0
U
14 E
4)
0
4n
a
rN o H~H - i
=1: =1~ =15:*
#x ## 4
=
=
H H-I
H- H-I
::= =1 =15: =t:
0
X#####X#4
40,
C)
dP
Ln
4-4
0
81
0
Uo
O0
.H
cn
C 0 J
-9
0ts
(w
rH
a04
en
U)
)l
'-
w
LLd
Or.
Q.U)
OcD
Oa)
O 0
a_
0J
ow
U)
Q.O
Uf)
0
,
LLJ
w
u
oO
O
O
C\J0
4-4
u ua
f~H
0
~.
0'14
0
0
-1/"IV =
0
·
NIVNiS
82
initiate compression.
This behavior is similar to the
results reported in the literature obtained with a spring
compression system (Emami, et al., 1979).
Another point
to be noted in figure 4-1 is that even though the temperature
set points are different for each experimental set, the degrees
of compression obtained for lower temperature settings (e.g.,
set #2 vs. set #5) are somewhat
higher.
The only apparent
variable parameter in the experiments shown in figure 4-1
was temperature and since it can be seen that the variation
of temperature in different sets does not have any particular
trend, the results in figure 4-1 were considered somewhat
surprising.
However, a closer examination of the initial
moisture content of the green beans used in these experiments indicated that a wide variation existed among different sets.
Figure 4-2 shows the same experimental results
after correcting moisture levels to an initial moisture content of about 900 g H 2 0/100 g solids.
It can be seen that
in figure 4-2 all the data points superimpose at corresponding levels of the applied pressures.
The fact that the
initial moisture content can affect the final degree of
compression at a given set of process conditions, relates
to the void volumes initially occupied by water.
For a given
sample it can be expected that the higher the initial
moisture content (and consequently a greater volume occupied
initially by water), the higher the degree of compression
at a given applied compressive pressure, since compression
83
.H -H
0
H -o
H
04-4
ro O
U>
o
o
o
0
N
r)
10
4.-
4-
4.-
0
pa
._
Ct
I
o Lr
0)
U)
d
d
0)
0)
0)
uw
>
O
0)
U)
C)
0n
.1>
0
LL
CL:
L
U)
0 >'
U)
o) Q
LL
0
En
o
o I
Cn,
4I
Oca
0r
Q
W U)
44 -H44
rT4
I
DO
d
I
I
(0
0
I / IV
0
I
I
?
''NlVtlS
84
of these void volumes does not require any force.
This in-
dicates that initial moisture content has to be taken into
account when compressing during freeze drying.
Figure 4-3 shows the degree of compression expressed
as compression ratio versus compressive pressure.
The
data points in this figure have been corrected for the
initial moisture content.
It can be noted that the com-
pression ratio increases linearly with compressive pressure
over the range of pressures tested.
This confirms earlier
results obtained for green beans and other products with
the spring compression system (Emami et al., 1979);when food
is under an applied compressive pressure during freeze
drying, it will undergo a deformation X until the resistive
force, -KX, is equal to the applied compressive force.
At this point the sample will undergo no further compression, which is similar to the behavior of a spring with a
constant modulus under stress.
Figure 4-4 shows the compression behavior as a
function of time during freeze drying at varying levels of
applied pressure.
It can be seen that the reduction in
volume has a uniform rate and is essentially linear with
time at a given applied compressive pressure; however, as
can be expected, the rate of compression increases as the
applied compressive pressure is increased.
By the end of
the process, compression levels off and reaches a final
value.
85
rl
0
)>
O
~4J
0 a)
cd
O m
d
ow.
o
0
U] d
Uma)
(D U
w
US
03
Cr)
W
o
D
't >
Wa)
w4,
t-Po
cn,
o
O
o
0
ff
0
I
('jcOo
E]
d
-I -H
4Q
IOH
-HH
N~
r()
It)
4-
4-
4-
C)
)
0n
)
a) 0
ra4
*0'
0
U0
O
I_
I
I
(0
v
O1iV
I
I
CQ
NOISS3HdVO3
I
0
86
Fig. 4-4.
Compression behavior of green beans as a function
of time during freeze drying at different
levels of compressive pressure.
0.8
i
/ ~-4 /4
/rA
A(
.
-J
II
0.6
/
A
Ar
zNJ
I-
C,
z
oZ
0
0.4
0
w
cL
w
0
LL
-J
> 0.2
0
·
7.5, 63%,0°C,20psi
A
7.5,63%,0°C,40psi
*
0
I
0
I
10
7.5,63%,0C,60psi
I
I
20
TIME (hours)
I
30
87
4.1.2
Dependence of the Drying Rates on Sample
Loading (Applied Compressive Pressure)
It was noted that compression during freeze drying
results in increased rates of drying which is presumably
due to improved heat transfer through the dry layer.
It
can be expected that there exists a limit to the improvement of the drying rate which can be achieved, since the
compression process should result in poorer mass transfer
through the dry layer.
Thus, at some level of energy input
and/or applied compressive pressure, the total process will
pass from heat transfer limited to mass transfer limited,
which can eventually result in collapse and irreversible
structural changes.
Figure 4-5 shows a typical example of
the effect of increasing the level of applied pressure on
drying rates.
It can be seen that the sample at the higher
level of the applied compressive pressure (60 psi) has higher
rates of drying at early stages of the process and reaches
the final moisture content in a much shorter time.
This
effect is even more significant when the drying rate of a
sample without any application of compressive pressure
(non-compressed) at the same conditions of freeze drying is
considered.
It was found that during the early stages of
the drying, the drying rates of compressed samples were as
much as twice as high as those of non-compressed samples (not
shown in the figure).
88
Fig. 4-5.
Effect of compressive pressure on drying rates of
green beans.
20
.
N
16
*00
-
3:
la
12
U)
w
i:
0
z
8
a
03
4
0
0
10
TIME
20
(hrs)
30
89
The sharp drop in the drying rates of the samples
being compressed during freeze drying is presumably due to
the smaller amounts of water left in sample, since most of
the water has been sublimated at the early stages of the
process.
This is indicated by lower drying times to reach
a given final moisture content at higher compressive pressures.
4.1.3
Relations of Extents of Drying and Compression
Relative to the Applied Compressive Pressure
When the data for compression of green beans during
freeze drying is analyzed in terms of volume reduction (percentage change of original height) and extent of drying
(percentage of the total water removed), if the removal of
a unit of water is associated with a unit fractional increment reduction in volume, a straight line relationship would
be expected.
The result of this analysis for various applied
pressures is shown in figure 4-6.
The data points are
given on an hourly basis up to 11 hours of compression/
freeze drying.
As the drying approaches completion, the
data points tend to become quite bunched since the changes
in volume and water content are very small.
For this reason,
data is not presented on an hourly basis after 11 hours.
The final degrees of compression
are shown at the 100%
water removal mark.
In figure 4-6 the effect of compressive pressures
of 20 to 60 psi are shown with other process
parameters
90
a)
-,-
Ct
0
w
rd
0
ro
a)
> 44
00
oo03
m
3~
U)
(3.
W
O
->)
r-I
a)
o4
,'"
00
W
:>
(3.
O
so
0)NU
U O
4-a)
tp
M.
0
0
0
3wn10^A NI NOIol3l
0
0
%
%4
91
fixed at 7.5 cm, 63%, and 0°C.
It can be seen that in-
creased compressive pressures improve both the extent of
drying and relative reduction in volume of the sample.
However, the increase in volume reduction with increasing
compressive pressure seems much greater than that of the
water removal, especially at later stages of drying.
It
is interesting to note that when the data points at each
hour are connected for the different compressive pressures
applied, a straight line is obtained for each hour as shown
in figure 4-6.
This indicates that the relative improvement
of the extent of drying and compression remains the same
as the compressive pressure is increased.
This effect, however,
seems more significant from 20 psi to 40 psi than 40 psi
to 60 psi.
The fact that the effect of the increased
applied pressure goes mainly into increasing the degree of
compression can also be seen from the change in the slope of
the straight lines at each hour.
4.1.4
Effect of Heat Input on the Compression
and Drying Behavior
Table 4-1 has the summary of the final degrees of
compression obtained at different levels of heat input.
It appears that for similar experimental conditions but
varying heating source temperatures (e.g., set #5; run #15,
10, and 18) where the controlling thermocouples were held
at 0°C, 100 C, and 200 C, respectively, there is little effect
on the final degree of compression.
92
Table 4-1 also shows the total drying time for the
same experiments being 24-25, 18-19, and 13-14 hours,
respectively.
The reason for the acceleration of the drying
rates due to increased heat input is obvious, as can be
seen in figure 4-7 (7.5 cm, 0%, 0°C, 40 psi vs. 7.5 cm,
0%, 20°C, 40 psi).
However,
close examination
of the rate
of compression at each temperature setting shows that there
is also an increase in the rate of compression at higher
temperatures.
Figure 4-8 shows the effect of increasing
the level of heat input on the extents of drying and compression.
It can be seen that all the data points essentially
fall on the same path (i.e., the relationship of water
removal and volume reduction is the same for all heat inputs).
However, at higher temperatures the relative change in both
water removal and volume reduction is greater at a given
time as indicated by the solid data points at selected
times in figure 4-8.
For instance, it can be noted that by
the end of the 10-hour period, the sample at 200 C has
reached final drying and compression, whereas the samples
at lower temperatures have not reached the end of the
process (figure 4-8).
This indicates that the effect of
increased heat input mainly accelerates the process,
without affecting the final degree of compression at a
given applied compressive pressure.
93
Fig. 4-7.
Effect of heat input, cell wall porosity, and
size on drying rates of green beans.
20C
-
V
(0
Z
>-
07
0
0
TIME
(hrs)
94
C
0
r
i-
-
O
w
>4
Crd
O
O0
<
0
O ) t
In
4oR
0
133
Q)
'-
0
0
4
o
oJ
w
C
>
>
0
4
0
i
CM
0 a
4 0 04
._
0
rd
a
0
o 0O
O
In
it
0O
-0
05
bo
0
o 4
I-
C 4ow
4- 04
4 Q)
U 4 D
O z U
PI tc ·
I
0
CD
I
I
0
I
I
0
Nu
3vyn-IOA NI NOlilOnatl3 %
I
0
0
"Ii
-I
jrx4
95
4.1.5
Effect of Size and Porosity on Compression
and Drying Behavior
In a series of experiments, the effect of the porosity
of the side walls of the sample cell holder (solid walls.
0% vs, 63% open area) was investigated.
From table 4-1 it
can be seen that there is very little effect of porosity
on the final degrees of compression (compare Run #2 and 15,
Run #1 and 10, or Run #1 and 19 in Set #5).
As can be expected, higher porosities of the cell
wall result in higher drying rates and shorter drying times
due to drying from the sides which is shown in figure 4-7
for otherprocess parameters fixed at 7.5 cm, 0°C, and 40 psi.
Similar to the results for porosity, no significant
effect was noted in the final degrees of compression for
two different sample holder sizes (7.5 x 7.5 cm2 vs. 10.5
x 10.5 cm2 ) .
Run #10 with
This can be seen in table 4-1 by comparing
#14, or Run #1 with #9, etc., in set #5.
Figure 4-7 shows the effect of sample holder size
on drying rates.
As mentioned earlier, the size of the
holders were chosen such that the larger sample holder contained twice as much green beans.
From figure 4-7 it can
be seen that the drying rates for the 10.5 x 10.5 cm2
sample holder remain about twice as much as for the 7.5
x 7.5 cm2 sample holder for process condition of 0%, 0° C
and 40 psi.
It can also be noted that the total drying
times to 5% final moisture content for both sizes shown in
figure 4-7 are not very different (24-25 hours).
96
Figure 4-9 shows the influence of porosity and sample
holder size on extents of drying and compression.
Again,
variation of these parameters does not appear to change
the general path (extent of compression vs. extent of drying)
independent of time.
4.2
Compression Behavior of Green Beans. Moisture
and Temperature Dependencies as Evaluated with
the Instron Universal Testing Machine
In the course of this research, it became apparent
that in order to characterize the compression behavior of
food materials during freeze drying, a more basic approach
was required, since studying compression behavior of foods
under freeze drying conditions is very complicated (i.e.,
interactions of moisture and temperature gradients and
other variables) and is also a very slow process (each data
point requires a complete freeze drying experiment under
controlled conditions).
It is apparent that deformation behavior of food
materials depends on their mechanical properties, the values
of which seem to be a strong function of moisture content
and temperature.
In order to determine the basic compressive mechanical
properties relevant to compression behavior during freeze
drying and the stress-strain behavior of this product as
a function of moisture and temperature, a series of compression studies were conducted with the Instron Universal
97
N
O
'4
'
F
64
&io
rO
,)
rdr
a)
rA
>o0
3
4
0
0
a
4
4
w
U
4
4
0
4
S
H-U)
00
._
.w
J
-)
CD
o
w
>
c-
4-)
U)
1U t0-0
G)>rQ.
._
4
a
0
00
rdH
*4 4
!
ItC*
P-
I .m
U
* 4
I
00
0
I
(D
0
3Jv4nOA
I
I
I
'4
NI NOI13AOn3
I
I
O
6O
%
O
O
&4
98
Testing Machine.
When mechanical stress is applied to food
materials (and other polymeric compounds), there exists
a critical range of temperature and moisture content for
which the solid structure can undergo deformation without
irreversible structural changes.
In this research a detailed
study of the influences of these variables (moisture and
temperature) on mechanical properties of green beans (freezedried initially) as related to compression of this product
was conducted.
The results were analyzed in terms of these
two variables (moisture and temperature) independently and
interdependently and further related to the mechanical
properties of this product.
4.2.1
Effect of Moisture and Temperature on
Compressibility of Freeze-Dried Green Beans
Freeze-dried green beans rehumidifed to different
moisture contents [up to about 45% moisture content
(dry basis)] and adjusted to different temperatures (+24°C
to -250 C) were compressed uniaxially in the Instron Universal
Testing Machine and the force-deformation relationship was
determined. From the force-deformation curves, the stresses
requied to achieve particular strains were determined for
samples of varying moisture contents and temperatures.
Figures 4-10 through 4-13 show the relationships of
applied stress and moisture content (Instron-derived,
stress-moisture curves) to achieve particular strains at
given sample temperatures with a cross-head speed of
99
n
a.
%l)
(I)
U)
0
10
20
30
40
% MOISTURE CONTENT (dry)
Fig. 4-10.
Effect of moisture
content on stress to
achieve given strains at +24°C (Instron
tests).
100
_
_
V
c
200
)
150
._
Cn
b
100
w
I-cr
C)
50
0
0
0
I
0
10
20
20
30
30
I
4
40
% MOISTURE CONTENT (dry)
Fig. 4-11.
Effect of moisture content on stress to
achieve given strains at +15°C (Instron
tests).
101
C
* 0.65
A 0.70
V 0.75
200 _
\*0.80 a
* 0.85
- 150
U
0.
V
b
CL
10
I00
w
C)
50
I
·IA
0
I
0
10
Ai
I
I
I
20
I
I
I
30
40
% MOISTURE CONTENT (dry)
Fig. 4-12.
Effect of moisture
content
on stress to
achieve given strains at -100°C (Instron
tests).
·
102
0
0.80
A s · 0.75
.r ln 7n
V
a a 0.65
._
cL
(V)
(I)
W
w
I-
Cl)
,,
F
0
I
I
10
20
I
30
I
40
% MOISTURE CONTENT (dry)
Fig. 4-13.
Effect of moisture content on stress to
achieve given strains at -25°C (Instron
tests).
103
20 mm/min.
Figure 4-14 shows the stress-moisture content
relationship for producing a strain of 0.75 for samples of
varying moisture contents and temperatures.
It can be seen that the general compressive behavior for green beans can be described as a decrease in
forces or stresses required to produce a fixed volume
reduction or strain as the moisture content increases,
until the moisture content reaches the range of 20%-35%,
at which point moisture content shows little further
influence.
The effect of temperature can be better visualized
in figure 4-14.
If the curve at -25°C is not considered,
the influence of temperature on the stresses to achieve a
given strain ( = 0.75 in this case) at a given moisture
content is similar (i.e., lower temperature requires more
stress), though the differences are less significant than
those for varying the moisture content.
It should be noted
that at -25°C most of water is in the solid state which makes
the material quite brittle.
This phase transition (water
to ice) can affect the validity of the stress-strain
relationship at this temperature relative to other temperatures tested.
The observed effects of increasing moisture content
and temperature on compressibility of the product are expected as described earlier in section 2.5 and can be
attributed to the decrease in relative rigidity of the food
structure.
104
200
-
O
- 25 0 C
A
- IO0C
o
- 15°C
V
- 240C
U)
QC,
C')
0
I
c)
0
A
\
0
10
C)
IA
V
I
20
|
30
|
% MOISTURE CONTENT (dry)
|
A
40
Fig. 4-14.
Effect of moisture content on stress to reach
a strain of 0.75 at different temperatures (Instron
tests)
105
4.2.2
Effect of Cross-Head Speed or Strain
Rate on Compressibility
In order to test the effect of the strain rate on
compresibility, freeze-dried, rehumidifed green beans at
various levels of moisture content were compressed uniaxially at room temperatures using the Instron Machine
at a cross-head speed of 50 mm/min.
Figure 4-15 shows the
stresses required to achieve given strains for both crosshead speeds of 20 mm/min and 50 mm/min at room temperature
(+240 C).
It can be noted that the results are very similar
and there is no apparent effect of the strain rate at
least for the two rates tested.
This would indicate that
at these two different levels of cross-head speed the
samples probably did not have any time for stress relaxation during the compression process that could result in
deviation in the force-deformation response.
As for the
compression during freeze drying experiments, although the
strain rates were very slow (1 cm/5- 10 hrs), however, due
to constant application of compressive pressure for the
duration of the experiment, it can be expected that the
compression of any region in the dry layer was instantaneous
with no time for stress relaxation.
4.2.3
Effect of Moisture and Temperature on
Mechanical Properties
Figures 4-16 through 4-18 illustrate the effect of
moisture and temperature on mechanical properties of green
106
e%
W+
I
I
z
.H
H
00
C
C0)
Z
t
,
o
U)
Uo
)
0
0O
U
(Isd)
0
O4 J
-0 'SS3UlS
00U)
,
--- ----
|
-
_
U)
-I rd
O
r)
I
W
V
* 4 Pu.,
U)
0r4
0)
0 a)
*
.:j~~~~~~~
40
yCYIYIY
4
D
a
U)
ro
a)
Q)
04
004)
0
W
*HI
00
0
-
m0
00
-Q)
0o
-
0
0
L1)
c
F
~R
ro
(T
Q)
0
l-
0
O
iI
Ii
0
ID
I
In)
O
( !sd)
D
'0
I·
O
'SS381.S
u
107
,ll
400
co
CL
a.
II
-
(v
,,)
200
3I0
z
Z
I-F
0
I
5
Fig. 4-16.
I
I
20
30
% MOISTURE CONTENT (dry)
10
!
40
Effect of moisture content (and temperature)
on tangent modulus calculated at strain
(£) = 0.65 (Instron tests).
108
_
_
7
i
400
a.
0
(/)
-J
0o
200
zbJ
I-
Z
0
!
I
5
10
I
20
I
l
30
40
I
% MOISTURE CONTENT (dry)
Fig. 4-17.
Effect of moisture content (and temperature)
on tangent modulus calculated at strain
(£) = 0.30 (Instron tests).
109
0
U)
4-
N.0O0
rd
a)
UQ
E
1: ow
0 -E
C-1
a,
·I
O
4
>r
·I
a D4
oz
O
r
0
Q) 4J)
L
Urd
11
r
(y)
d4CH
0O
i
0-
O
rE
co>
I
0ao
I
I
I
0'
1
I
0
1
I
0
!sd Iz a D0 sn-naon INV3S
Hl
0
0
.LNV
H
110
beans.
The mechanical properties which were measured are
tangent modulus (a/c) calculated at two different strains
(0.65 and 0.30) (figures 4-16 and 4-17 respectively), and
secant modulus (a/E) calculated at a stress of 21 psi
(figure 4-18).
As expected, it can be noted that the effect
of increaseing both moisture and temperature causes a
reduction in the values of these mechanical properties both
of which represent relative rigidity of the material.
4.2.4
Moisture-Temperature Dependencies: Common
Mechanical Properties and Similar Compression
Behavior (Stress-Strain Relationship)
The relative effects of moisture and temperature
on compressibility of freeze-dried, rehumidifed green beans
suggest the existence of specific combinations of these two
variables which will result in.commonvalues for mechanical
properties of the food and in a similar compression response
under compressive stress.
In fact, similar stress-strain relationships are
obtained for freeze-dried green beans having specific combinations of moisture and temperature which correspond to
common values of the mechanical properties investigated.
Figures 4-19 through 4-21 present the moisture-temperature
relationships for specified values of tangent modulus and
secant modulus which give similar stress-strain relationship for freeze-dried green beans.
It should be noted
that each individual curve from the family of curves
1ll
O
Wr
w
I-
10
20
30
40
% MOISTURE CONTENT (dry)
Fig. 4-19.
Moisture-temperature relationship with
equivalent compression behavior at specified
values of tangent modulus (at e = 0.65) obtained in Instron tests.
112
U
0
-
w
5
10
20
30
40
% MOISTURE CONTENT (dry)
Fig. 4-20.
Moisture-temperature relationship with
equivalent compression behavior at specified
= 0.30) obvalues of tangent modulus (at
tests.
Instron
in
tained
113
0I
W
n-
W
O.
W
I-
% MOISTURE CONTENT (dry)
Fig. 4-21.
Moisture-temperature relationship with
equivalent compression behavior at specified
values of secant modulus (at a = 21 psi)
obtained in Instron tests.
114
presented corresponds to a specified value of a given
mechanical property, and further all the moisture-temperature
combinations which fall on each individual curve (i.e., have
a specified mechanical property) have similar compression
behavior (stress-strain relationship).
Based on the results presented in Figures 4-10
through 4-13 and figures 4-19 through 4-21, it is possible
to calculate compressive behavior of freeze-dried green
beans which have been rehumidifed and adjusted to any combination of temperature and moisture content.
Given either
a value of moisture and temperature or a value of a pertinent
mechanical property, it is possible to predict and construct
a stress-strain relationship for freeze-dried green beans.
This is described further in section 4.3.
4.3
Correlation of Compression Behavior during
Freeze Drying to Compression Behavior with the
Instron Universal Testing Machine
4.3.1
Moisture-Temperature Relationships and
Mechanical Properties of the Dry Layer
during Freeze Drying
According to the results presented in section 4.2
and literature information, it can be stated that for a
given food material, the mechanical properties at any location in the dry layer of the freeze-drying sample will depend
on the local temperature and moisture content.
Considering
moisture and temperature gradients that occur in the dry
layer during freeze drying,it can be expected that there
115
exist various combinations of moisture and temperature
which will give mechanical properties suitable for compression during freeze drying.
If it is assumed that the same factors which govern
compressive behavior of the food in the Instron Universal
Testing Machine also apply to the compressive behavior during
freeze drying, it is then possible to relate the two processes.
To achieve this, it is first necessary to consider the temperatures and moistures occurring in freeze drying.
As reported earlier (section 2.8), Gentzler and
Schmidt (1973) showed that if the temperature of the
frozen region during freeze drying process remains constant,
then at any instant in time, the temperature and moisture
content of any location in the dry layer zone are fixed.
Furthermore, for each given temperature in the dry layer,
there will be a unique value of moisture content at that
location whose value can be determined from figure 2-11
for skim milk and granulated coffee.
The estimated moisture
and temperature relationship in the dry layer for green beans
during freeze drying at a constant ice region temperature
of -250 C is shown in figure 4-22.
It was shown earlier that for any given value of
moisture and temperature, there will be a fixed value of
each mechanical property, whose value for tangent modulus
and secant modulus can be estimated from figure 4-19 through
4-21.
Thus, given the moisture-temperature relationship
116
0
N r-
rd)
O
)
0
-r-I]
Al:4-
O
O
LLJ
rd Q 0
(1) u
()1 *H Cd
LLI
H
Oa
Or
4
LL-
U
m- ,l
H
Urd
k rdOJ
LLO
4e
U)
a40J
o4 0
04
O
0
0
(sp!los 6001/5)
O
C
831VM 331-NON
4-
--I
-
Ul
117
that exists in the dry layer during freeze drying (figure 4-22),
it is possible to calculate the range of the values of the
given mechanical property in the dry layer during freeze
drying.
As an example, figure 4-23 shows how this can be
done for tangent modulus at
= 0.65 by combining figures 4-19
and figure 4-22 and assuming an ice-front temperature.
In
figure 4-23 fixed values of temperatures with corresponding
values of moisture contents in the dry layer during freeze
drying (from figure 4-22) are transferred to figure 4-19
which represents moisture-temperature relationship at specified
values of tangent modulus (=0.65) obtained in Instron
studies.
Combination of these two provides the values of
the given mechanical property at each location of the dry
layer corresponding to those local moistures and temperatures.
By knowledge of the values of the mechanical property at
each location in the dry layer, it is possible to construct
the range of the mechanical property as a function of dry
layer temperature or distance from the ice interface which
is shown in the bottom of figure 4-23 for tangent modulus
at
=
0.65.
Figure 4-24 and 4-25 show the estimated
range of tangent modulus at
= 0.30 and secant modulus at
a = 21 psi as a function of dry layer distance.
It should
be noted that in these figures location of the minima,
which is close to but not adjacent to the ice-front, indicates the most compressible region that exists in the dry
layer at any instant in time.
This implies that if a constant
118
_ 20
o I0
II
u
w
-
I
.
2
2
0
z
JI
I
0
I
I
I
!
II
20
40
60
TEMPERATUREDIFFERENCE(C)
Ist.td
frr-dryr
oaUl
_-aur.
ot
loal
9ro
.
br.
La.-ron-
a.ta
Ln
n t.
day I·yr
*a fro.on
diffa.-r.
taraur
of. -2S-C
OISTURE CONTENT11ry)
duin
h..n
'iature
bU.:t!t=
bahaao
otiad
-U-.
t.
at apItfiad
1a. Iuaa.u
.tahp
.. iwa
o
aih
t-..
qalta iaat
.dalia
(at
.a.-U
t.Et(
//
V
a
(I
42
w
A.
- 600
% MOISTURECONTENT
(dry)
oof 500
/
Z 400
z
I
14
-10
-20
0
10
DRY LAYER TEMPERATURE,
20
C
OC ISTANCE
Iann. Ut t
fta. a.. d-
Fig. 4-23.
.,ot ue
Utrl -y-0
I...O.$(l)
la ,,i-
tUt
aut
1
tLM dy
lyr.
durl
Estimated values of tangent modulus (at
= 0.65) in the dry layer during freeze drying
of green beans as a function
dry layer.
of distance
in the
119
-cn
0
o
Cd
mq4
Od
o40
Et
(
UIO
)
I\0
CLLLa
)
0
0Co
oc
U0)
J)UC
~
W(
O
>u-I.,
rd
H En
. -H U
I
0(0
00
O'o
I
t0It.
· 'snlnnaow
I
":r
I
0
NQ
IN39Ntl.
0
.01
120
-
I
4J
r
UO
U)
0
CM
0
II -*o
40-
0
2Lj
o
N
)
m
Lii
w
LU
z
ro
a-
UC
0
8
rd
O
3
I-
Cd
)
c
0
I
I
O
a3
I
0
(D
rl )1 FO
M
)
W Q) I
UN
0'
J
I
O
qv*
!sd IZ = D 'Ssn-ncoi
I
I
0N~
O
Idd
.1-1
INVO3S
121
compressive pressure is applied to the sample at any instant in time during freeze drying, each location in the dry
layer will undergo a deformation (compression), the degree
of which depends on the value of the mechanical property of
that location relative to the applied compressive pressure.
Furthermore, the highest degree of compression will occur
in the location with minimum value of modulus.
Figure 4-26 shows the behavior of the modulus as
a function of time and distance.
Assuming a constant ice-
front temperature means that the value of the minima does
not change with time; however, its location moves inward
as the ice front recedes, except for the initial and final
stages of the drying process.
Initially the drying surfaces
are very cold resulting in high values of the modulus (high
rigidity).
At the end of the drying process when ice dis-
appears and the moisture gradients drop, the minima also
disappear resulting in a rigid structure again.
It is
interesting to note that the rotation of the gradient that
occurs about the constant surface temperature produces
decreasing values for the temperature gradient, and it is
thus expected that the distance of the minima from the icefront does not stay constant as drying proceeds.
shown in figure 4-26.
This is
122
I
C
ow
ri r
ord4
o.C
o.
*
_= otL
1..
.
4- 4
o'
,o
000
0
E-° °'
._ *-s "
l_E "v
-UO
W
U~~~
= -m
-H >,
Caa
r- N
E aW
.
-H d
C
.1
0
-H
>
=(
W
a)
a) -H a
N
Cr'
123
4.3.2
Mechanism of Compression during Freeze
Drying at Constant Applied Compressive
Pressure
It was shown earlier that from knowledge of the
value of the given mechanical property and the applied compressive pressure, the degree of compression that will occur
can be determined.
Thus considering the behavior of the
modulus in the dry layer at any instant in time at a constant applied compressive pressure, we note that the dry
layer will undergo a certain deformation, the degree of
which can be determined at each location in the dry layer
from the value of the modulus at that location and time.
As mentioned before, the maximum compression will always
occur at the location corresponding to minimum rigidity in
the dry layer, and thus, the minimum value of the modulus.
It can be expected that, since the minima moves
inward through the dry layer with.time (figure 4-26) with
a constant value, all the locations in the dry layer will
eventually undergo a fixed maximum compression which
depends on the minimum value of the modulus, and the
given applied compressive pressure.
This will result in
a fixed compression throughout the sample by the end of the
compression/freeze drying process.
4.3.3
Development of Means to Predict Compression
Behavior during Freeze Drying; Compressive
Pressure-Strain Relationship
Considering that if moisture and temperature of an
124
already freeze-dried sample are independently adjusted so
that the sample has a value of the modulus which corresponds to the minimum of the modulus that exists in the
dry layer during freeze drying, it can be expected that both
samples will have equivalent compression behavior for the
same applied stress.
In other words, the previously freeze-
dried sample with adjusted uniform temperature and moisture
content will have the same mechanical properties as that
location in the dry layer of the freeze drying sample
having minimum value of the modulus.
In fact when compressive pressure-strain data for
compression during freeze drying (figure 4-2) are transferred to the Instron-derived, stress-moisture curves for
freeze-dried samples with adjusted uniform moistures and
temperatures (figures 4-10 through 4-12), it is noted that
for each Instron test temperature the compression during
freeze drying data points fall on a straight line passing
through a single moisture content as shown in figure 4-27.
This means that from the Instron-derived, stress-moisture
curves, a single moisture content at that particular Instron
test temperature (in this case +150 C) actually simulates
the compressive pressure-strain relationship (figure 4-2)
obtained for a series of compression during freeze-drying
experiments.
To construct figure 4-27, specific compressive
pressure-strain values were taken from figure 4-2 and were
reconstructed on the Instron-derived, stress-moisture curve
125
--
i
i
4
A
rn.
vI3,
V
mt2,
VI
-4·C
I
4·C
t 4,
o'C
0.11
0
20
40
60
80
COMPRESSIVE PRESSURE (psi)
Ufet
c
·
,urstr*r
P
4zyl
l
-lam
_a iotr
f - -
_
p
% MOISTURE CONTENT(dry)
Wf.tI
t
· t a =1rlt c
ctUl
-
t
tU.
"=U
rUU --
.u.l
i
IUb
V
·
200
0.70
on
.
o.
150
% MOISTURECONTENT(djr)
b
; 100
Lj
0:
I-
I
\\\a|
cy
*
T
_6
)_
50
%,
O
10
20
30
--L
40
% MOISTURECONTENT(dry)
Fig. 4-27.
Representation of the equivalent moisture content
simulating compression during freeze drying data
at Instron test temperature of +15°C.
126
at 150 C (figure 4-11) to match the exact values of the
pressures (stresses) and strains.
It can be seen that by
connecting the resulting points, a straight line is obtained
passing through a single moisture content (14.5%)
on the
Instron-derived, stress-moisture curve at +15°C which is
shown at the bottom of figure 4-27.
This can also be
done with the Instron-derived, stress-moisture curves obtained
at other temperatures as shown in figure 4-28 for the Instron
test temperatures of +24°C and -100 C.
However, the equivalent
moisture contents that simulate compression during freezedrying data (figure 4-2) at these other Instron test temperatures have changed values.
This change in equivalent
moisture content relative to the Instron test temperature
is expected, and it merely reflects the fact that at a
given value of the mechanical property, there are other
specific combinations of moisture and temperature which
result in similar compression behavior.
Table 4-2 gives
the specific combinations of moisture contents and temperatures, with corresponding values of the mechanical properties
which simulate the compression behavior of green beans during
freeze drying as a function of the applied compressive
pressure (results obtained in figure 4-2).
In other
words, if moisture contents and temperatures of already
freeze-dried green beans are adjusted to these specific
combinations shown in table 4-2 or correspond to the values
of the mechanical properties shown in this table, their
127
U)
U)
4rd
a
-
4-o
z
o
z
.p
U)
4.+
-.
Cr
nrl o4
Q)
W 4
U]
%16
o o
(!sd) 0 'SS381SS
40
U)
4
G)H
~4o
a)
t
z
H N
00
CM
4JWQ)
(Z
Q)
O
0
c
CN
0.
&~
(!sd) D'SS311S
-I-I
12 8
tQ *l
0l
-H
rl
aO U)
C
o
0
aU
ON
II
CU3
tnO
o,
*H C
-H
a
04U),
-4Cd
U)
00
4-'-
o
4 0
D CU
Cd
o U)
U
tn
3U)
Uo
-,U)
-H
U)-H
En
70
u
11
O)
u
0,H
= d
0cd
U)
kFU)Vl
z
U)
Erq
UOd
4CU)o
E
a
o
0'
o
N
0
o
"
+
+
E d
U)
,-
m
OU)
-I
H
r
I
4-3
'-4
OP
k
,
N
129
stress-strain relationship as obtained in the Instron
Universal Testing Machine will simulate the compressive
pressure-strain data obtained in compression during freezedrying experiments (as shown in figure 4-2).
In order to evaluate compressive pressure-strain
dependency of the food in compression during freeze drying
(figure 4-2), sevearl experiments have to be conducted.
Each data point in figure 4-2 corresponds to one experiment
in which the food is compressed during freeze drying at a
constant applied compressive pressure which takes about
10-24 hours.
However, rapid and simple compression tests
with the Instron Universal Testing Machine (1.5 min per test)
also can determine complete compressive pressure (stress)strain relationship of the food with a specified value of
a mechanical property.
Since complete information on the values of the moduli
that exist in the dry layer during freeze drying are not
available, an additional compression during freeze-drying
experiment is necessary in order to use this predictive
method.
Once a compressive pressure-strain data point is
established for compression during freeze drying, the whole
compressive pressure-strain relationship can be predicted
from stress-moisture relationships which can easily be
obtained from Instron studies.
The errors in compression behavior that are involved in this predictive method for +1% deviation in the
130
estimation of the equivalent moisture content for the
Instron tests at +240C are shown in table 4-3 and figure 4-29.
It can be seen that in figure 4-29 these deviations are
reasonable and within the experimental errors.
4.3.4
Evaluation of the Results Obtained in Compression
during Freeze-Drying Experiments Relative to the
Parameters Tested Based on the Results Obtained
in Instron Studies
Using the correlations of Instron and freeze-drying
studies which are presented above, the influence of various
process parameters on compression behavior during freeze
drying can be discussed in more depth.
Table 4-4 gives a
summary of the overall effects of these process variables
examined in compression during freeze-drying studies.
4.3.4.1 Effect of the Applied Compressive Pressure
on the Drying Rates and the Degree of Compression
As shown in table 4-4 it was noted that increasing
the applied pressure acts to increase the degree of compression and to accelerate the drying process.
The latter
is attributed to improved thermal conductivity and heat
transfer through the already dried/compressed layer.
Com-
pression process during freeze drying eliminates or decreases
the void volumes between individual pieces and internal voids
that are produced by sublimation of ice crystals.
This
results in better contact between and within the food solids
and consequently a better heat transfer through the dried/
compressed layer.
131
Table 4-3
Variations in the stresses to reach a given strain
caused by
1% variation
in the estimation
of the
equivalent moisture content at 240 C in the Instron
tests
Strain
Stress (psi)
0.65
27+3
0.70
36+5
0.75
53+7
0.80
90+15
132
-
9
0
O _-~
a
>
\O
H rH
o+1~
o
II
I :
0
,°
I l:
o
o
o
d
-
C
C
o 0
0
4rd
0
0
-g
a:0
,
0
I .
I
>
OC
LU
+
0Ld
.
H *rl ·
0o
I
0 o.
U)
I
*
1:
U)
r-O
flTn
Q)NO
a-H
Ua
I04 n orI _
U) 4Q.
o
- M
a)
oHd c
m 4-H0f
0 r-I 4a)
2
1~--)C
c
k U)l >Hf
-
1I
0
1I
-
I
o
tS1
1/1V
= 'NI
(5o
.\\
I
I
o
c
rl4
.H
133
Table 4-4
The effects of parameters tested in compression during freeze drying on final degree of
compression and drying time
Freeze Drying
Variable
Final Degree
of Compression
Drying Time
Applied compressive
pressure
Large
Some
Control temperature
Not Significant
Large
Porosity of sample
holder
Not Significant
Slight
Size of sample
holder
Not Significant
None
(heat input)
134
Provided that the ice temperature remains constant,
improved heat transfer results in acceleration of the icefront movement.
It was shown earlier (figure 4-26) that the
rate of movement of the minimum of the modulus is a function
of the rate of the ice-front movement.
Thus acceleration
in the ice-front movement increases the rate of movement of
the minimum of the modulus.
As a result of these effects,
faster drying and faster and higher degrees of compression
can be expected when the applied compressive pressure is
increased (figure 4-6).
From figure 4-6 it can be seen that
these effects were observed [i.e., increasing the applied
compressive pressure acted to increase both the water
removal (i.e., faster drying time) and the degree of compression at given times].
4.3.4.2 Effect of Increasing Levels of Heat Input
on Drying Rates and the Degree of Compression
It is clear that increasing the level of heat input
accelerates the freeze-drying process.
However, within the
limits tested, it did not show any apparent effect on the
final degrees of compression (figure 4-8).
This can be
explained by noting that, under the condition that the temperature of the ice region remains constant, the moisturetemperature relationship which exists in the dry layer does
not change with incresaed temperature gradients across the
dry layer, associated with the increased levels of heat input.
This explanation is based on results obtained by Gentzler
135
03'
0.
--0---
_
=t
0
00<
04
0
ro
0
tn
>- 4O
0
,.,.
A
o-
o 0
a)E
0
e
C
W
k >
'O0
e
1.
C
rd
3Z
to
I,
O
e.
-C
Iv
0
Vr
0O
;
-)
._
._
ffi
ff
CL
O
O
O
C
it
I,:
O
I
0
O0
4
C
I
0
O
0
I
3Jfn'10A
I
0'4.
I
I
0
NI NOI.Lna3dlG %
I
0
0
C
(1)
g4M
P4 LH
O
I
Q)
136
and Schmidt (1973) which indicate that as long as the
temperature of the frozen region during the freeze drying
process remains constant, then the temperature and moisture
contents in the dry layer zone are interrelated in such a
way that at any instant in time, each temperature value in
the dry layer has a unique corresponding moisture content.
Furthermore, as long as the ice-front temperature remains
contant, the unique temperature-moisture content relationship does not depend on location in the dry layer or time
of drying.
Measuring a given temperature value at a loca-
tion in the "dry" layer simultaneously determines the moisture
content at that location.
Similarly, increased temperature
gradients associated with incresased levels of heat input
do not change the temperature-moisture relationship across
the dry layer, provided that the frozen region temperature
does not change in value.
This results in the same value
of the minima of the modulus (i.e., the same relative
rigidity) in the dry layer, at different levels of heat
input (i.e., different constant surface temperatures).
This behavior is illustrated in figure 4-30.
It can be seen
that at a fixed ice-front temperature, the fact that the
moisture-temperature relationship in the dry layer remains
constant at different surface temperatures (i.e., different
levels of heat input) means that the value of the minima
of the modulus does not change, resulting in the same
compressibility in the dry layer for a given applied
137
tangent modulus (TM1 ) at Ts1
tangent modulus (TM2) at T2
E
I
I
DRY
I
LAYER
T
/
S2
TsS,
DISTANCE
Fig. 4-30.
Effect of increasing the surface temperature
(Ts) during freeze drying on the range of
modulus in the dry layer at constant ice-front
temperature.
138
compressive pressure. Since increased levels of heat input
accelerate the drying process and consequently the rate of
the movement of the ice-front, it can be expected that the
rate of the movement of the minima of the modulus will also
increase.
This results in a shorter process time to reach
the same degree of compression and final moisture content
as the level of heat input is increased (figure 4-8).
As
shown in figure 4-8, it can be seen that all the data points
for different levels of heat inputs essentially fall on
the same path (i.e., the relationship of water removal and
volume reduction is the same for all heat inputs).
However,
at higher surface temperatures, the relative change in
water removal and volume reduction is greater at any given
time.
For instance, as shown by the solid data points in
figure 4-8, it can be seen that by the end of the 10-hour
period the sample at +20°C has already reached the final
drying and compression, whereas the samples at lower surface
temperatures have not reached the end of the process and
require more time to reach to the same values of the final
moisture content and final degree of compression.
4.3.4.3 Effect of Size and Porosity of the Sample
Holders on Drying and Compression Behavior
As mentioned before, the size and porosity of the
sample cell holders used in these experiments did not seem
to have any significant effects on drying and compression
behavior of green beans, and thus there is no reason to
139
.H
O)
~ ,
_
0jrd
O
ro
a)
o
r
rk
O o
(O
* .('
o~
o s At
O
W
i
C
>
C12
H 0
o
tH
0) N
-
0a)
0
0
0
0
3(nqOA N NOLlOnC3
N
3nnIOA
NOIiL3f3U
%
0
°/
O
2Z4
140
believe that these variables significantly alter the behavior of the modulus in the dry layer during freeze drying.
4.4
General Discussion
Compression of foods during freeze drying can be
considered as an efficient method of producing high-density,
low-moisture products.
This technique minimizes the number
of processing steps that are required to produce freezedried/compressed foods.
It has been shown how moisture and temperature
gradients can influence the mechanical properties of the
solid structure in the dry layer such that compression can
be accomplished during freeze drying.
Interrelationships
of compression and freeze-drying processes were investigated,
and the effects of some parameters that can influence compression and drying behavior of green beans were evaluated.
The independent effects of moisture and temperature
on compression behavior of the food (initially freeze-dried)
were investigated with an objective method (i.e., Instron
Universal Testing Machine).
A literature investigation
showed that despite the important influence of these two
variables on physical and chemical properties of food
systems, there has been very little information reported
in the area of rheology and mechanical properties of food
porducts,.especially for low-temperature and moisture levels.
141
The results of the above studies were further related
to compression during freeze drying to demonstrate that the
combination effects of the same factors are responsible for
the compression behavior in both processes.
A simple method was developed in order to utilize
the compression data obtained with the objective method for
the prediction of the compression behavior of the foods
during freeze drying.
Figure 4-26 illustrates the relationship between
moisture-temperature gradients and the modulus as a function
of time in the dry layer during the freeze-drying process.
It can be seen that the minima of the modulus corresponds
to a specific combination of moisture and temperature.
Furthermore, the value of the modulus will stay constant
so long as the frozen region temperature does not change
irregardless of the surface temperature, or size and
geometry of the sample.
When there is a constant applied
compressive pressure on the sample during the freeze-drying
process,all the locations in the dry layer will undergo a
deformation, the value of which depends on the local value
of the modulus, with maximum compression taking place where
the minima of the modulus is located.
Considering the be-
havior of the minima with respect to time, it can be seen
from figure 4-26 that the minima of the modulus eventually
passes through all the locations in the dry layer.
Thus,
application of a constant compressive pressure will result
142
in uniform deformation of the dry layer eventually resulting
in a fixed degree
of compression
at the end of the process
whose value is determined by the applied pressure and the
value of the minimum of the modulus.
As an example, con-
sider the compression behavior at locations A, B, and C
in the "sample" shown in figure 4-31 at various times during
the freeze-drying process. As long as these locations are
in the frozen region, no compression will take place as
shown in the figure at time = to .
At t1 the ice-front has
just passed the location A, so that there will be some compression at that location, the extent of which. depends on
the value of the modulus at location A and the constant
pressure being applied to the sample.
At t2 location A
has already reached its maximum degree of compression, since
the minima has just passed through the location A.
At time
t 3 it can be seen that the value of the modulus at location A
has increased, which means the relative rigidity at that
location is higher than at time, t2 , and thus there will be
no further compression at the given applied compressive
pressure.
However, at t 3 the ice-front has just passed
through location B, so that some compression will start to
take place at that location.
At t4 location B is com-
pressed to its maximum degree, which is the same degree of
compression as that of location A at time t2 , since the
value of the minima and the applied compressive pressure
have not changed.
At t5 the relative rigidity of the
143
Ici
-H
so
4O-
.-
4-
-0
4-
0o
0)
rd
,44
ro
0
,1 N
Qrd
O
-
Co
M
d
l) O
H >1
.H
14 4
location B has;increased so no further compression will
take place there; meanwhile (at t5) the ice-front has
just passed through location C, and compression will start
to take place at this location until the maximum degree of
compression is reached at time t6.
Considering that all
the locations in the dry layer will eventually go through
this cycle, the end result will be a uniform and fixed
degree of compression throughout the sample by the end of
the process.
It should be noted, however, that at the very
beginning and the very end of the process, compression
behavior can be expected to be slightly different, due to
the two end effects.
Initially the drying surfaces are very
cold, which results in a high value of the modulus as shown
in figure 4-31 at time to.
This results in greater rigidity
of the sample surface and consequently lower degree of compression at the surface.
At the end of the drying, when ice
disappears and moisture gradients drop, the minima will also
disappear, resulting in a rigid structure again (figure 4-31,
at t9 ).
The above gives the general mechanism of the compression behavior during the freeze-drying process.
The
knowledge of the value of the minima of the modulus and the
applied compressive pressure will give information on the
final degree of compression.
Since there is not enough
information on the values of the mechanical properties of the
145
dry layer during freeze drying to theoretically calculate
compression behavior, the simple predictive method which was
explained in detail in section 4.3.3 can be used to predict
the compression behavior of foods during freeze drying.
146
5.
CONCLUSIONS
The results and conclusions of this study may be
summarized
A.
as follows:
Study of mechanical properties (Instron studies)
1.
Increases in moisture content above the B.E.T.
monolayer value and up to about 25% (dry solid basis)
decrease the rigidity of freeze-dried green beans which have
been rehumidified.
2.
The freeze-dried product, rehumidified up to the
moisture content of approximately 25% can be compressed with
no apparent structural damage.
The compressed product
partially recovers upon unloading and has high recovery upon
rehydration.
3.
Increasing temperature from -25°C to +24°C in
the freeze-dired product having moisture contents of approximately 6%-30% also reduces the relative rigidity.
However, the effect of temperature is less significant than
that of moisture.
4.
For given values of the mechanical properties
there are specific combinations of moisture and temperature
which result in equivalent compression behavior (i.e.,
similar stress-strain relationship).
Therefore, the compres-
sive behavior can be predicted at any given combination of
moisture and temperature.
147
5.
Based on the information in the literature
it can be deduced that for each temperature
in the dry layer
during freeze drying there is a unique moisture content, and
for each combination of temperature and moisture there is
a corresponding value of each mechanical property.
On this
basis, by consideration of the moisture and temperature
gradients that exist in the dry layer during freeze drying,
it can be stated that at each instant in time, each location in the dry layer will have a different mechanical
property (i.e., different compressibility characteristics).
This results in a range of mechanical properties throughout
the dry layer which change value at each location as the
ice-front recedes with time.
Furthermore, as long as
the temperature of the ice-front remains constant, the value
of the minimum modulus will not change.
6.
It can be expected that application of com-
pressive pressure at any instant in time during freeze
drying will cause compression of the dry layer.
The degree
of compression, however, at each location will depend on
the stress-strain relationship at that location.
Further-
more, the maximum compression will take place at locations
which have lowest values of the modulus (minima).
7.
Considering that the fixed value of the minimum
modulus (rigidity) will eventually pass through all locations in the dry layer during freeze drying, at a constant
applied compressive pressure, all locations in the dry
148
layer will eventually be compressed to the same ultimate
degree.
B.
According to the results obtained in the compression
during freeze drying experiments and part A, it can
be stated that:
1.
The main process parameter that affects the
degree of compression during freeze drying is the level of
the applied compressive pressure.
2.
Application of compressive pressure during
freeze drying also increases drying rates.
This is pre-
sumably due to improving heat transfer (and in particular
the thermal conductivity) through the already dried/compressed
layer.
3.
Higher heat inputs to accelerate the freeze drying
process result in faster movement of the minimum modulus
and, therefore, in faster compression.
However, when the
temperature of the ice--frontremains the same, similar final
degrees of compression will be obtained for a given compressive pressure, at different heat inputs.
4.
Size and porosity of the sample holders (used
in this study) which do not change the ice-front temperature will not affect the minimum modulus; thus, similar
degrees of compression will be obtained at the given
applied compressive pressure.
149
6.
FUTURE RESEARCH
The research reported upon here has described and
explained the compression behavior of green beans, both
under iso-moisture/iso-thermal conditions (Instron studies)
and under freeze drying.
A more complete and wide-ranging understanding of
compression behavior of foods at low moistures and/or
temperatures and during freeze drying can be achieved by
the future research in the following areas.
1.
The independent and interactive effects of
moisture and temperature on mechanical properties of
other freeze-dried food materials using the Instron
Universal Testing Machine.
This area of research relates to the variation of
the basic material structure and its effects on compression
behavior.
It can be expected that the composition and
organization of the structural polymers (such as carbohydratebased polymers in plants and protein-based polymers in
animal tissue), and those low-molecular-weight compounds
that are located between the polymeric segments result in
different mechanical properties which can be affected
differently by moisture and temperature.
It is thus im-
portant to investigate the influences of moisture and
temperature on compression behavior of other foods to demonstrate the existence of specific combinations of these
150
two variables which result in common values for mechanical
properties and in a similar compression response of the
given food (irregardless of the basic material structure)
under stress.
2.
To investigate the effects of moisture and
temperature on mechanical properties of a model system or
artificial food structures.
The advantage of using a model
system simulating foods is that it is easy to achieve major
changes of the composition in order to evaluate the effects
of the basic structural components in the model system.
3.
A more detailed study of the conditions which
result in collapse and irreversible structural changes due
to compression.
Although some tests were conducted to
evaluate the effect of the level of heat input on collapse,
further studies are required to collect sufficient information in order to be able to predict the conditions which
result in irreversible structural changes in compression
during freeze drying.
4.
Evaluation of the microstructural changes that
occur due to compression at different levels of moisture
content and temperature and various basic material structure,
using microscopic techniques.
Rahman et al. (1969) reported
that the most damage to the cellular structure of freeze-dried/
compressed carrots was due to the freezing process prior
to dehydration and compression and further that compression
151
of the product caused no significant damage to the cellular
structure.
However, further work on microstructural
changes due to compression would be of interest especially
with respect to collapsed systems.
This area of research
relates to the evaluation of the microstructural changes that
occur at various conditions which result in collapse and
irreversible structural changes in compression of food
materials.
5.
Determination of moisture and temperature gradients
that exist in the dry layer of various foods during freeze
drying (espeically moisture levels close to ice-front) at
different conditions of heat and mass transfer.
This will
give the knowledge of the mechanical properties and thus
compressibility of the dry layer using the data obtained
for the same foods from the Instron studies.
This would
allow prediction of the compression behavior during freeze
drying.
6.
To model a mathematical simulation of freeze
drying which includes changes in food physical properties
due to compression.
It should be realized that due to the
complex nature of the process, a number of simplifying
assumptions will probably have to be made.
The mathematical
model can be used to predict the time dependence and
overall compression with freeze drying by incorporating
the compression information obtained from the Instron studies
152
directly into the model.
This will eliminate the need
for actual freeze drying experiments for the prediction
procedure.
The overall methodology will follow directly
from those used in the mathematical simulation, an iterative
procedure over small time intervals in which heat and/or
mass transfer balances would be used to evaluate temperature and moisture gradients, water loss, and interface
movement.
The mechanical properties for each time interval,
which are a function of moisture and temperature at each
location, will be obtained from the information obtained
with Instron tests.
The values of the mechanical properties
would be combined with the value of the applied force to
calculate the extent of compression at each location.
Changes in food physical properties due to the compression
would be made prior to the next iteration.
153
BIBLIOGRAPHY
*Aguilera,
(1974).
J. M. and J. M. Flink.
"Technical
Note:
Determination of moisture profiles from temperature
measurements during freeze drying," J. Fd. Technol. 9,
391-396.
Beke, G. (1969). "The effect of sublimation temperature
on the rate of freeze drying process and upon the
volumetric change in meat muscle tissue," In: Proceedings of the XII International
tion, Vol. 3, Madrid.
*Brajnikov,
Congress
A. M., A. I. Vassiliev,
of Refrigera-
V. A. Voskoboinikov,
and E. I. Kantchechivili. (1969). "Transfer de
chaleur et de masse dans les materiaux poreaux pendant
le lyophilisation sous vide," In: Symposium on Thermodynamic Aspects of Freeze Drying, International Institute
of Refrigeration, Commission X, Lausanne, Switzerland.
*Bralsford,
R.
(1967).
J. Fd. Technol.
2, 339-353.
*Brockman, M. C. (1966). "Compression of foods," U.S. Army
Natick Laboratories Activities Report, 18 (2), 173-177.
*Do, J. Y., C. Srisangnam,
D. K. Salunkhe,
and A. R. Rahman.
"Freeze-dehydrated and compressed sour cherries,
(1975).
I, Production and general quality evaluation," J. Fd.
Technol. 10, 191-201.
*Donnelly, J. C. (1947). "Methods of processing foodstuffs,
debulking food for packaging," U.S. Patent 243,334,8.
*Drake, B. (1966). "Advances in the determination of texture
and consistency of foodstuffs," In: Proceedings of 2nd
International Congress of Food Science and Technology,
Warsaw, Poland.
Duckworth, R. B. (1972). "The properties of water around
the surfaces of food colloids," Proc. Inst. Food Sci.
Technol.
(U.K.) 5 (2) , 60.
*Durst, J. R. (1967).
"Compressed
food components
to
minimize storage space," Technical Report 68-22-FL,
U.S. Army Natick Laboratories, Natick, Mass.,AD-662060.
*Emami, S-H. (1976). M.S. Thesis.
of Technology, Cambridge, Mass.
Massachusetts Institute
154
*Emami, S-H., J. M. Flink, and A. R. Rahman (1979). "Compression of foods during freeze drying," J. Fd. Processing
and Preservation 2 (4), 285-298.
*Ferry, J. D.
(1961).
Viscoelastic
2nd ed., p. 518, New York:
Properties
of Polymers,
John Wiley & Sons, Inc.
*Gentzler, G. L., and F. W. Schmidt. (1973). "Thermodynamic
properties of various water phases relative to freeze
drying," Trans. ASAE 16, 179-182.
*Gooding,
E. G. B., R. B. Duckworth,
D. C. MacDougall,
and
J. Gasay. (1958). "The storage life of dehydrated
vegetables III," Unclassified Report SA/DA/RM/82, Ministry
of Agriculture, Fisheries and Food, Great Britain.
*Hamdy, M. M. (1960). "Compression of dehydrated foods,"
Review of Literature, Contract No. DA19-129-QM-1630,
Quartermaster Food and Container Institute for the
Armed Forces, Chicago, Illinois.
*Hamdy, M. M. (1962). "Compression of dehydrated foods,"
Final Report, Contract No. DA19-129-QM-1899, Quartermaster Food and Container Institute for the Armed
Forces, Chicago, Illinois.
*Haralampu, S. G. (1977). M.S. Thesis, Massachusetts
Institute of Technology, Cambridge, Mass.
Hardin, T. C. (1965). Ph.D. Thesis, Georgia Institute of
Technology.
Hatcher, J. D.
Technology.
(1964). M.S. Thesis, Georgia Institute of
*Ishler, N. I. (1965). "Methods for controlling fragmentation of dried foods during compression," Technical Report
D-13, U.S. Army Natick Laboratories, Natcik, Mass.
*Jones,
R. L. and G. J. King.
(1975).
"Use of hydrating
salts to accomplish limited freeze drying," For presentation at symposium on Drying: Principles and Technology,
National Meeting, American Institute of Chemical
Engineers, Boston, Mass.
*Kapsalis, J. G., J. E. Walker, Jr., and M. Wolf. (1970).
"A physico-chemical study of mechanical properties
of low and intermediate moisture foods," J. Texture
Studies 1, 464-483.
155
*Karel, M. (1974).
In: Advances in
Foods, A. Spicer
Publishers, LTD,
"Fundamentals of Dehydration Processes,"
Preconcentration and Dehydration of
(editor), pp. 44-94, Applied Science
London
*Karel, M. (1975). Freeze Dehydration of Foods. In:
O. R. Fennema (editor), Principles of Food Science,
Part II,'Physical Principles of Food Preservation, Vol. 4,
pp. 359-395. New York: Marcel Dekker, Inc.
Kessler, H. G.
(1962). Chem.-Ing. Tech. 34:163.
*King, G. (1950). "The swelling of textile materials,"
J.S.D.C. 66:27-34.
*King, C. J., W. K. Lam, and 0. C. Sandall.
Technol. 22, 1302.
(1968).
Food
*King, C. J. (1971). Freeze Drying of Foods," CRC Press.
Cleveland, Ohio.
*King, C. J., R. M. Carn, and R. L. Jones (1975). "Processing approaches for limited freeze drying," For presentation at annual meeting, Institute of Food Technologists,
Chicago, Illinois.
*Konigsbacher, R. S. (1974). "Methods for storing shape
and structure of compressed dehydrated animal and combination products," Technical Report 75-56-FL, U.S.
Army Natick Laboratories, Natick, Mass.
*Labuza, T. P., K. Acott, S. R. Tatini, R. Y. Lee, J. M.
Flink, and W. McCall. (1976). "Water activity determination: A collaborative study of different methods,"
J. Fd. Sci. 41 (4), 910-917.
*Longmore, P. (1973).
Industry, 970.
"Freeze drying," Chemistry and
Luikov, A. V. and L. L. Vasiliev. (1969). "Heat transfer
of capillary porous bodies in a rarefied gas flow,"
In: Symposium on Thermodynamics Aspects of Freeze
Drying, International Institute of Refrgeration,
Commission X, Lausanne, Switzerland.
*Luyet, B. J. (1962). In: F. R. Fisher (editor), FreezeDrying of Foods, National Academy of Science - National
Research Council, pp. 194-211, Washington, D.C.
*MacKenzie, A. P. and B. J. Luyet. (1964). "Apparatus
for freeze drying at very low, controlled temperatures,"
Biodynamic 9 (183), 177-191.
156
MacKenzie, A. P. (1965). "Basic principles of freeze
drying for pharmaceuticals," Bull. Par. Drug Assn.
20 (4), 101.
MacKenzie, A. P.
101.
*MacKenzie,
(1966). Bull. Parenteral Drug Assn. 20,
A. P., and B. J. Luyet.
(1969).
"Recovery
of
compressed dehydrated foods," Technical Report FL-90,
U.S. Army Natick Laboratories, Natick, Mass. AD-708901.
*MacKenzie,
A. P. and B. J. Luyet.
(1971).
"Recovery
of
compressed dehydrated foods, Phase II," Technical
Report 72-33-FL, U.S. Army Natick Laboratories, Natick,
Mass.
*Margaritis, A., and C. J. King. (1969). Paper presented
at A.I.Ch.E. Meeting, Portland, Oregon.
Meffert,
H. F. T.
(1963).
"Heat transmission
freeze
drying material," Proceedings of the XI International
Congress on Refrigeration, Munich.
*Meffert, H. F. T.
(1965).
"Freeze drying:
Physical
aspects," IBVT (now Sprenger Inst.) Ann. Report,
P.63, Wageningen, The Netherlands.
*Mohsenin, N. N. (1968). Physical Properties of Plant and
Animal Materials, Part I, Vol. 1, First preliminary
edition. Department of Agricultural Engineering,
Pennsylvania State University.
*Nissan, A. H.
(1976). "H-bond dissociation
in hydrogen
bond dominated solids," Macromolecules 9 (5): 840-850.
Oetjen, G. W. (1973). "Continuous freeze drying of granulates
with drying times in the 5-10 minutes range," Proceedings
of XII International Congress
Avi, Washington, D.C.
*Pavey, R. 0.
(1975).
of Refrigeration
"Study techniques
3, p. 697.
for controlling
flavor intensity in compressed foods," Phase I, Technical
Report 75-49-FEL, U.S. Army Natick Laboratories, Natick,
Mass.
*Peleg, M.
(1977).
"The role of the specimen
dimensions
in uniaxial compression of food materials," J. Fd.
Science
42
(3) : 649-651.
157
*Pilsworth, M. N., Jr. and H. J. Hoge. (1973). "The
compression of freeze-dried beef to form bars; plasticizing with water transferred as a vapor," Technical
Report 73-56-PR, U.S. Army Natick Laboratories, Natick,
Mass.
*Rahman, A. R., G. Schafer, G. R. Taylor, and D. E. Westcott.
(1969). "Studies on reversible compression of dehydrated
vegetables," Technical Report 70-36-FL, U.S. Army Natick
Laboratories,Natick, Mass., AD-698457.
*Rahman, A. R., G. R. Taylor, G. Schafer, and D. E. Westcott.
(1970). "Studies of reversible compression of freezedried RTP cherries and blueberries," Technical Report
70-52-FL, U.S. Army Natick Laboratories, Natick, Mass.
*Rahman, A. R., S. Bishov, and D. E. Westcott, (1971).
"Reversible compression of dehydrated peas," J. Texture
Studies 2, 240-245.
*Ranadive, A. S., E. A. Huang, and E. Seltzer. (1974).
"Investigation of rehydration characteristics of compressed comminuted meats," Technical Report 74-53-FEL,
U.S. Army Natick Laboratories, Natick, Mass.
*Sakata, I. and R. Senjue. (1975). "Thermoplastic behavior
of lignin with various synthetic plasticizers," J. Appl.
Polymer Sci. 19, 2799-2810.
Sandall, O. C. (1966). Ph.D. Thesis, University of
California, Berkeley.
Sandall, O. C., C. J. King, and C. R. Wilke. (1967). "The
relationship between transport properties and rates of
freeze drying of poultry meat," A.I.Ch.E.J. 13, 428.
*U.S.D.A. (1948). "Experimental compression of dehyrated
foods," Miscellaneous Publication 647, presented by
Agricultural Research Administration, Washington, D.C.
*Literature reviewed